The longest US routes

What is the longest US route?

That question seems like it should be pretty simple.  And if you ask an internet search engine, most results will show the same, simple answer.  However, that answer is incorrect, and the mileage provided is overstated (as I will demonstrate below).

It actually required a surprising amount of research in order to obtain the correct answer.  I have shared the results below, not only for the longest US route currently, but also for the longest US routes historically.

For those who may not want to wade through all of the data/research/analysis that follows, I'll make it easy by providing the answers here at the top of the page:

• US 30 was the longest route from 1926-1931.

• US 50 was the longest route from 1931-1937.

• US 6 has been the longest route from 1937 through the present (although a complexity has existed since 1963, requiring us to differentiate between continuous routes and discontinuous routes):

  • 1937-1963: US 6 was the longest route.

  • 1963-2019: US 6 was the longest continuous route, but the combined total of US 20[e] and US 20[w] was longer.

  • 2019-present: US 6 is the longest route, once again longer than even the combined total of US 20[e] and US 20[w].

US 30 was the longest route from 1926-1931

1927 log

• 3472* (US 30, Atlantic City NJ to Astoria OR)

• 3220 (US 40, Atlantic City NJ to Oakland CA)

• 2856 (US 50, Annapolis MD to Sacramento CA, but initially there was a gap between Thistle UT and Ely NV, so this was a combined total of US 50[e] from Thistle to Annapolis plus US 50[w] from Ely to Sacramento.)

• 2726 (US 80, Savannah GA to San Diego CA)

• 2542 (US 20e, Boston MA to Yellowstone Nat'l. Park WY)

• 2481 (US 10, Detroit MI to Seattle WA)

• 2448 (US 66, Chicago IL to Los Angeles CA)

• 2330 (US 70, Beaufort NC to Holbrook AZ)

• 2269 (US 1, Ft. Kent ME to Miami FL)

• 1382 (US 60, Virginia Beach VA to Springfield MO)

• 1301 (US 12, Detroit MI to Miles City MT)

• 766 (US 64, Conway AR to Capulin NM)

• 707 (US 6, Provincetown MA to Erie PA, but initially there was a gap between Danbury CT and Kingston NY, so this was a combined total of US 6[e] from Provincetown to Danbury plus US 6[w] from Kingston to Erie.)

*The figure of 3472 miles was almost certainly overstated, but it is not possible to analyze it because the 1927 log listed only the cumulative national total for each route (not the state-by-state or point-to-point totals).  Therefore anyone with access to this log would have naturally concluded that US 30 was the longest route, and they would not have had any reason to question the figure of 3472 miles, at least not until the next edition of AASHO's route log was published in 1929.

1929 log

The 1929 edition did include point-to-point mileages, and in analyzing those I believe US 30's figure of 3308 miles was plausible.  That figure was 164 miles less than the one listed in 1927, which calls into question the accuracy of the 1927 figure.  It is reasonable to assume that the true mileage of US 30 wouldn't have changed much in just two years, so in 1927 it was probably closer to 3308 (rather than 3472).  But even if so, that still would have put its mileage well above that of US 40.  For that reason I am reasonably confident in stating that US 30 was the longest route from 1926 to 1931.

• 3308 (US 30, Atlantic City to Astoria)

• 3259 (US 50, Annapolis to Sacramento, with no gap)

• 3212 (US 40, Atlantic City to Oakland)

• 2676 (US 80, Tybee Island GA to San Diego)

• 2505 (US 20e, Boston to Yellowstone)

• 2499 (US 66, Chicago to Los Angeles)

• 2440 (US 10, Detroit to Seattle)

• 2330 (US 70, Beaufort to Holbrook)

• 2290 (US 1, Ft. Kent to Miami)

• 1427 (US 12, Detroit to Miles City)

• 1291 (US 60, Virginia Beach to Springfield)

• 879 (US 64, Conway to Raton NM)

• 735 (US 6, Provincetown to Erie, with no gap)

For someone using the 1929 log, it would not have been a simple matter to determine the longest US route, because US 30, 40, and 50 all had split routes, and the log's method of listing mileages made it necessary to manually add some figures together in order to come up with a national cumulative total for each route.

US 50 was the longest route from 1931-1937

In 1931, US 50 was extended from Sacramento west to Oakland, which pushed its mileage barely above that of US 30.  This was reflected in AASHO's next route log, the 1932 edition, although this change would not have been readily apparent to the casual user, because manual addition was still necessary in order to obtain the national cumulative total for any highway that included split routes (and that remained the case in the '35 and '38 logs as well).

1932 log

• 3291 (US 50, Annapolis to Oakland)

• 3243 (US 30, Atlantic City to Astoria)

• 3163 (US 40, Atlantic City to Oakland)

• 3162 (US 60, Virginia Beach to Los Angeles)

• 2636 (US 80, Tybee Island to San Diego)

• 2424 (US 20e, Boston to Yellowstone)

• 2417 (US 66, Chicago to Los Angeles)

• 2400 (US 10, Detroit to Seattle)

• 2331 (US 62, Niagara Falls NY to El Paso TX)

• 2289 (US 1, Ft. Kent to Miami)

• 2240 (US 6, Provincetown to Greeley CO)

• 2193 (US 64, Ft. Landing NC to Santa Fe NM)

• 2190 (US 70, Atlantic NC to El Paso TX)

• 1455 (US 12, Detroit to Miles City)

1935 log

• 3287 (US 50, Annapolis to San Francisco)

• 3243 (US 30, Atlantic City to Astoria)

• 3163 (US 40, Atlantic City to San Francisco)

• 3122 (US 60, Virginia Beach to Los Angeles)

• 3085 (US 70, Atlantic to Los Angeles)

• 2636 (US 80, Tybee Island to San Diego)

• 2424 (US 20e, Boston to Yellowstone)

• 2417 (US 66, Chicago to Los Angeles)

• 2397 (US 10, Detroit to Seattle)

• 2331 (US 62, Niagara Falls to El Paso)

• 2289 (US 1, Ft. Kent to Miami)

• 2274 (US 6, Provincetown to Greeley)

• 2193 (US 64, Ft. Landing to Santa Fe)

• 1455 (US 12, Detroit to Miles City)

US 6 has been the longest route from 1937 through the present

(although a complexity was introduced in 1963)

In 1937 US 6 was extended from Greeley to Long Beach, increasing its mileage to significantly more than that of US 50.  The next route log (the 1938 edition) indicated 3652 miles for US 6, which was the longest distance ever listed for any route in any of AASH(T)O's logs.

1938 log

• 3652 (US 6, Provincetown to Long Beach CA)

• 3295 (US 50, Annapolis to San Francisco)

• 3223 (US 30, Atlantic City to Astoria)

• 3178 (US 40, Atlantic City to San Francisco)

• 3117 (US 60, Virginia Beach to Los Angeles)

• 3022 (US 70, Atlantic to Los Angeles)

• 2605 (US 80, Tybee Island to San Diego)

• 2447 (US 20e, Boston to Yellowstone)

• 2397 (US 10, Detroit to Seattle)

• 2351 (US 66, Chicago to Santa Monica CA, with major reroute in NM)

• 2324 (US 62, Niagara Falls to El Paso)

• 2280 (US 1, Ft. Kent to Miami)

• 2193 (US 64, Ft. Landing to Santa Fe)

• 1744 (US 12, Detroit to Yellowstone)

In 1940 a separate western segment of US 20 was commissioned.  US 20[w] has never connected to US 20[e], so adding the mileages of the two segments together yields a misleading figure, because it ignores the approx. 70-mile gap between the two segments.  But it would also be misleading to include those ~70 miles in the total, because that mileage passes through Yellowstone National Park, and neither AASHO nor the National Parks Service has ever considered any road in Yellowstone to be a part of US 20.  So the most accurate way to think of US 20 is not as a single route, but rather as two discontinuous routes that share the same number (more about this on the Gaps page).

Starting with the first route log after US 20[w] was commissioned (the 1942 edition), AASH(T)O has always acknowledged the gap between the two segments.  But despite that, the '42, '51, and '55 logs each did what I described above as misleading: they combined the mileage of the two disconnected segments of US 20.  Nevertheless, for the sake of comparison, I have included the combined mileage of the two US 20s in the lists below.  If one were to accept that combined total of "US 20" as a valid measurement, then it became the second-longest route, because it barely displaced US 50 from that position.  However, US 6 was still significantly longer than the combined "US 20".

1942 log

• 3576 (US 6, Provincetown to Long Beach)

• 3277 (combined total of US 20e and US 20w)

• 3244 (US 50, Annapolis to San Francisco)

• 3138 (US 40, Atlantic City to San Francisco)

• 3093 (US 30, Atlantic City to Astoria)

• 3017 (US 60, Virginia Beach to Los Angeles)

• 2933 (US 70, Atlantic to Los Angeles)

• 2568 (US 80, Tybee Island to San Diego)

• 2446 (US 1, Ft. Kent to Key West FL)

• 2418 (US 20e, Boston to Yellowstone)

• 2346 (US 62, Niagara Falls to El Paso)

• 2309 (US 10, Detroit to Seattle)

• 2285 (US 66, Chicago to Santa Monica)

• 2195 (US 64, Ft. Landing to Santa Fe)

• 1702 (US 12, Detroit to Yellowstone)

• 859 (US 20w, West Yellowstone MT to Albany OR)

The 1942 log was the first to include the mileage of the shortest split route in the cumulative total for any route that had a split, making it a simple matter to find the true end-to-end mileage for any route.  This was the case in the '51 and '55 logs as well.

1951 log

• 3533 (US 6, Provincetown to Long Beach)

• 3319 (combined total of US 20e and US 20w)

• 3316 (US 50, Ocean City MD to San Francisco)

• 3167 (US 30, Atlantic City to Astoria)

• 3150 (US 40, Atlantic City to San Francisco)

• 3004 (US 60, Virginia Beach to Los Angeles)

• 2926 (US 70, Atlantic to Los Angeles)

• 2568 (US 80, Tybee Island to San Diego)

• 2422 (US 1, Ft. Kent to Key West)

• 2401 (US 20e, Boston to Yellowstone)

• 2288 (US 62, Niagara Falls to El Paso)

• 2283 (US 10, Detroit to Seattle)

• 2271 (US 66, Chicago to Santa Monica)

• 2174 (US 64, Nags Head NC to Santa Fe)

• 1685 (US 12, Detroit to Yellowstone)

• 918 (US 20w, West Yellowstone to Newport OR)

1955 log

• 3517 (US 6, Provincetown to Long Beach)

• 3300 (combined total of US 20e and US 20w)

• 3231 (US 50, Ocean City to San Francisco, but no longer via Salt Lake)

• 3150 (US 30, Atlantic City to Astoria)

• 3093 (US 40, Atlantic City to San Francisco)

• 2963 (US 60, Virginia Beach to Los Angeles)

• 2915 (US 70, Atlantic to Los Angeles)

• 2568 (US 80, Tybee Island to San Diego)

• 2418 (US 1, Ft. Kent to Key West)

• 2401 (US 20e, Boston to Yellowstone)

• 2283 (US 62, Niagara Falls to El Paso)

• 2277 (US 10, Detroit to Seattle)

• 2257 (US 66, Chicago to Santa Monica)

• 2179 (US 64, Nags Head to Santa Fe)

• 1713 (US 12, Detroit to Yellowstone)

• 899 (US 20w, West Yellowstone to Newport)

Things get more complicated

On June 19, 1963, AASHO approved California's requests to truncate the two longest continuous US routes (plus several others).  The US 6 designation was removed from the segment between Long Beach and Bishop CA.  This would have caused US 50 to replace US 6 as the longest US route, were it not for the fact that on the same day, the US 50 designation was removed from the segment between San Francisco and Sacramento.  Thus, despite its truncation, US 6 retained its distinction as the longest continuous US route.  However, the mileage of US 6 had dropped below that of the combined total of the two US 20s.  This would not have been common knowledge immediately, because the next route log published after these changes wasn't until 1968, and even then it listed only state-by-state mileage totals.  This made it necessary to do some addition in order to come up with the cumulative national totals.

1968 log

• 3247 (combined total of US 20e and US 20w)

• 3202 (US 6, Provincetown to Bishop)

• 3089 (US 30, Atlantic City to Astoria)

• 3073 (US 50, Ocean City to Sacramento)

• 2875 (US 40, Atlantic City to Truckee CA)

• 2701 (US 60, Virginia Beach to Blythe CA)

• 2619 (US 70, Atlantic to Blythe CA)

• 2492 (US 12, Detroit to Aberdeen WA)

• 2374 (US 80, Tybee Island to Yuma AZ)

• 2364 (US 20e, Boston to Yellowstone)

• 2361 (US 1, Ft. Kent to Key West)

• 2253 (US 62, Niagara Falls to El Paso)

• 2242 (US 10, Detroit to Seattle)

• 2127 (US 64, Nags Head to Santa Fe)

• 1888 (US 66, Chicago to Needles CA)

• 883 (US 20w, West Yellowstone to Newport)

Like the '68 log, the '74, '79, and '89 logs also listed only state-by-state totals, not the national cumulative total.  Nevertheless, it was still a relatively simple matter to add up the state-by-state totals for a route in order to figure out its national total.  However, starting with the 1974 log, AASHTO made things even more confusing.  Towards the front of the book they printed a Table of Contents page with three columns: the US route number, the page on which it was found, and its "mileage".  So why bother adding up the state-by-state subtotals, when the overall mileage was listed right there?  Well, let's take a look at the "mileages" of some selected routes on the TOC page of the 1974 log:

• 3366 (US 20)

• 3355 (US 50)

• 3234 (US 6)

• 3183 (US 30)

• 2926 (US 40)

Now, if we compare those figures to the mileages in the previous log (the '68 edition, above), we find that every route had an increase in mileage, some by significant amounts:

• US 20 "increased" by 119 miles

• US 50 "increased" by 282 miles

• US 6 "increased" by 32 miles

• US 30 "increased" by 94 miles

• US 40 "increased" by 51 miles

Note that none of the terminus points for any of those routes had changed since '68.  So, did the mileage of all those routes really increase?  Of course not.  But then, what was behind all of those discrepancies?  Well, AASHTO provided an answer right there on the top of the TOC page:

In other words, the mileages on the TOC page did not represent the end-to-end mileage of the mainline US route.  Rather, those figures represented the combined mileage of the mainline US route plus the mileages of any associated auxiliary routes.

As an example, let's look at the mileages listed in the detail section for US 20 in order to see how AASHTO came up with a figure of 3366 miles:

2371: mileage of US 20[e]

881: mileage of US 20[w]

84: mileage of Alternate US 20[e] in NY

5: mileage of Business US 20 in Elgin IL

21: mileage of Business US 20 in Rockford IL

4: mileage of Business US 20 in Idaho Falls

3366: Total

Why did AASHTO do it that way?  My assumption is that they were less concerned with the end-to-end mileages of the mainline US routes, and more interested in the overall mileage of the entire US route network.  By lumping in the mileage of all the auxiliary routes under their respective mainline route, they could then add all the mileages listed on the TOC page, and the result would be the total mileage of the US route network (including all auxiliary routes).

That's fine, but unfortunately many researchers over the years have either failed to read AASHTO's statement at the top of the TOC page, or else the significance of that statement has eluded them.  What it means, again, is that the mileage listed on the TOC page cannot be assumed to represent the end-to-end mileage of the mainline US route.  For any US route that has one or more auxiliary routes, one must instead add the state-by-state subtotals in order to arrive at the correct end-to-end mileage.  But too many researchers have not done that.  Instead they have simply looked at the figures on the TOC page and mistaken them for end-to-end mileages.  Similar to the disclaimer above, here are the ones that were included with the '79 and '89 logs:

Below I have provided the actual end-to-end mileages for the longest mainline routes in each of AASHTO's last three route logs ('74, '79, and '89):

1974 log

• 3252 (combined total of US 20e and US 20w)

• 3203 (US 6, Provincetown to Bishop)

• 3115 (US 30, Atlantic City to Astoria)

• 3059 (US 50, Ocean City to Sacramento)

• 2874 (US 40, Atlantic City to Truckee)

• 2683 (US 60, Virginia Beach to Blythe)

• 2489 (US 12, Detroit to Aberdeen)

• 2376 (US 80, Tybee Island to Yuma)

• 2371 (US 20e, Boston to Yellowstone)

• 2368 (US 70, Atlantic to Globe AZ)

• 2363 (US 1, Ft. Kent to Key West)

• 2269 (US 64, Nags Head to Farmington NM)

• 2245 (US 62, Niagara Falls to El Paso)

• 1973 (US 10, Detroit to Spokane WA)

• 1333 (US 66, Joplin MO to Needles)

• 881 (US 20w, West Yellowstone to Newport)

1979 log

• 3229 (combined total of US 20e and US 20w)

• 3196 (US 6, Provincetown to Bishop)

• 3105 (US 30, Atlantic City to Astoria)

• 3008 (US 50, Ocean City to Sacramento, via I-70)

• 2688 (US 60, Virginia Beach to Blythe CA)

• 2488 (US 12, Detroit to Aberdeen WA)

• 2368 (US 20e, Boston to Yellowstone)

• 2364 (US 1, Ft. Kent to Key West)

• 2363 (US 70, Atlantic to Globe)

• 2301 (US 40, Atlantic City to Park City UT)

• 2247 (US 62, Niagara Falls to El Paso)

• 2244 (US 64, Nags Head to Farmington)

• 2012 (US 80, Tybee Island to Benson AZ)

• 1785 (US 10, Detroit to Missoula MT)

• 1162 (US 66, Joplin to Sanders AZ)

• 861 (US 20w, West Yellowstone to Newport)

1989 log

• 3237 (combined total of US 20e and US 20w)

• 3207 (US 6, Provincetown to Bishop)

• 3073 (US 30, Atlantic City to Astoria)

• 3011 (US 50, Ocean City to Sacramento)

• 2670 (US 60, Virginia Beach to Brenda AZ)

• 2483 (US 12, Detroit to Aberdeen)

• 2385 (US 70, Atlantic to Globe)

• 2377 (US 1, Ft. Kent to Key West)

• 2376 (US 20e, Boston to Yellowstone)

• 2326 (US 64, Nags Head to Teec Nos Pos AZ)

• 2298 (US 40, Atlantic City to Park City)

• 2248 (US 62, Niagara Falls to El Paso)

• 1700 (US 80, Tybee to Anthony NM)

• 861 (US 20w, West Yellowstone to Newport)

• 706 (US 10, Bay City MI to West Fargo ND)

Post-1989 measurements

The 1989 route log was the last one that AASHTO compiled, so at the time of this writing it was almost four decades out of date.  Yet today many researchers act as if the 1989 log is still accurate, throwing around words like "official" and pretending that nothing has changed during all those years.  Among the most frustrating examples are a pair of reciprocal mileage signs installed in 2016 at both endpoints of the combined "US 20".  These are very cool signs (and I wish other DOTs would do something similar) except for the fact that the mileage was significantly overstated:

The 3365 mile figure came from the TOC in the 1989 route log, which (as I explained above) includes not only the mileage of US 20[e] and US 20[w], but also the mileage of three US 20 business loops and a US 20 alternate route.  So that figure was overstated by 128 miles.

Additionally, even if the actual 1989 mileage had been used (3237 miles), nearly 30 years had passed by the time the DOTs consulted the route log in 2016.  So it was a misstep to use that log as a reference, because the fact is: many things have changed since 1989.  New highway alignments are built frequently, and each one of those changes affects the overall mileage of the route.  Following is a partial list of changes to both US 6 and US 20 which took place after 1989:

• 2004: the US 20[e] freeway opened between Ft. Dodge and Waterloo IA, reducing the distance by 19 miles.

• 2013: the US 20[e] expressway was completed between Ft. Dodge and Sioux City IA, reducing distance by 6 miles.

• 2016: the US 20[w] bypass was completed around Eddyville OR, reducing distance by 3 miles.

• 2019: US 6 was changed to bypass downtown Council Bluffs, increasing distance by 3 miles.

Those changes would have been enough to reduce the distance along the combined "US 20" by 28 miles, for a total of 3209.  US 6, meanwhile, would have increased to 3210.  So, based only on those changes, we would expect US 6 to be about one mile longer than the combined "US 20" now.  (And indeed, something very close to that difference is borne out in more recent measurements, as we will see below.)  However, since it is likely that additional changes to US 6 and US 20 (besides those listed above) have taken place in other parts of the country since 1989, the question becomes: how can we get accurate, current mileage figures?  It seems to me there are two reasonable options:

• Obtain the mileage figures from each state DOT and add them together.

• Use Google Maps (or similar) to create driving directions and record the mileage.

I have used both methods, and I have found that each one has its own benefits, as well as downsides.  So I actually recommend using both, because that way each method can be used to reality-check the other.  Below I discuss in more detail my experiences and observations.

The online map method

Following are the mileages I calculated in 2020 using Google Maps:

• 3205.4 (US 6, Provincetown to Bishop)

• 3203.3 (combined total of US 20e and US 20w)

• 3078.5 (US 30, Atlantic City to Astoria)

• 3015.0 (US 50, Ocean City to Sacramento)

• 2654 (US 60, Virginia Beach to Brenda)

• 2477.7 (US 12, Detroit to Aberdeen)

• 2370 (US 1, Ft. Kent to Key West)

• 2365.6 (US 70, Atlantic to Globe)

• 2346.5 (US 20e, Boston to Yellowstone)

• 2332.9 (US 64, Nags Head to Teec Nos Pos)

• 2297.4 (US 40, Atlantic City to Park City)

• 2246.9 (US 62, Niagara Falls to El Paso)

• 1030.4 (US 80, Tybee Island to Dallas)

• 856.8 (US 20w, West Yellowstone to Newport)

• 713.3 (US 10, Bay City to West Fargo)

Following are my observations about using the online map method:

Pros:

• Mileage can be obtained for all states, even those with uncooperative DOTs.

• Using the same platform to measure mileages in all states seems more likely to yield an apples-to-apples comparison.

Cons:

• Google driving directions tend to diverge from the US route and instead use a nearby freeway, so care must be taken to "pin" the driving route to the actual highway being measured.

• Potentially, Google could label a particular road as a US route, but that route may not be the actual US highway as defined by the authoritative DOT.  So one must make sure they are measuring the correct, current US route, even if Google does not display it that way.

• If the US route being measured happens to be under construction, Google sometimes disallows the use of that road in its driving directions.

• Google's distances are precise only to the nearest tenth-mile (which seems like it should be adequate, but in some cases it may not be).

For the above reasons and more, some people do not consider Google Maps to be a reliable source for mileages.  In order to address those concerns, and in order to find out if the mileage of US 6 had truly re-surpassed that of the combined "US 20", I next turned to the DOT method.

The DOT method

Historically the state DOTs provided AASHTO with mileage data to publish in their route logs.  Even though AASHTO is no longer doing that, it is still possible (theoretically) to use current DOT figures in order to calculate updated mileages for any US route.  Following are the distances I calculated in 2022 for US 6 and US 20 by cobbling together mileages provided by the state DOTs:

• 3216.379 (US 6, Provincetown to Bishop)

• 3204.766 (combined total of US 20e and US 20w)

What specific DOT sources were used?

Here is a link to the Google Sheet that I use to maintain up-to-date totals of both sets of mileages (i.e. the mileage according to Google Maps, and the mileage according to the various state DOTs).  I also use that sheet to keep track of specifically where I obtained the DOT info, so included there are links to DOT online sources (if applicable), as well as several tabs that contain screenshots of DOT-provided information.  A close examination of that sheet might raise the following questions:

Why is there PennDOT mileage for US 20, but not US 6?

Actually, there is a PennDOT source that reports statewide mileage for both US 6 and US 20.  However, the mileage for US 6 was over-reported by about 40 miles, so I did not cite that figure.  I have contacted PennDOT about this more than once.  On one occasion a PennDOT employee divulged that their figure was overstated because it included the mileage of not only mainline US 6, but also US 6N, plus three US 6 business routes (in other words, the same problem as AASHTO's route log).  But my followup attempt to get a breakdown of the mileage for each of those individual routes was not successful.

Why don't you have the mileages through Illinois?

IDOT publishes lots of their data online, but unfortunately their data that lists mileage by route number is limited only to interstate highways.  I have attempted to contact IDOT multiple times, but have never received a response.

What about Colorado?

CDOT makes their Online Transportation Information System (OTIS) available to the public.  This system shows that the US 6 milepost at the Nebraska line is 464.287.  However, that figure cannot be used for US 6's statewide mileage, because there are a couple places where CDOT's mileposts do not correspond to actual road mileage (as is common in any state where a highway has been realigned after the mileposts were placed).  These offsets result in the mileposts understating the total length of US 6 by about seven miles.  I contacted CDOT to obtain more accurate statewide mileage, but did not get a response.

Pros: 

• Most people consider DOTs to be an official and reliable source (although in reality it is absolutely necessary to fact-check their mileages rather than simply accepting them at face value).

• DOTs usually have very precise measurements, down to the hundredth- or thousandth-mile (which may seem like overkill, but when comparing two routes that are as close in mileage as US 6 and US 20, those tiny fractions of a mile could potentially make a difference).

Cons: 

• Not all DOTs provide their mileage figures online.

• Among those that do, the figures can be unusable for various reasons.  A couple recurring examples: 

  • figures include the mileage of auxiliary routes (i.e. the same problem as AASHTO's most recent route logs).

  • figures are based on mileposts (which usually do not get re-set whenever the route gets realigned).

• Not all DOTs respond to requests for (or questions about) mileages.

Conclusion

Current measurements using both methods (online maps and DOT mileages) agree that the end-to-end mileage of US 6 is longer than the combined total of US 20[e] and US 20[w].  So not only has US 6 been the longest continuous US route since 1937, but it is also once again longer than any discontinuous US route, including US 20.

Research and/or image credits: Wayne Brunelle; Mike Roberson; Dale Sanderson; Michael Summa; Dick Tobiason