Longest east-west stretches along normally north-south highways, and vice versa

[Dirt Roads]

WV 4, which parallels I-79 and US 119, is a diagonal, but covers 2.5x as much east-west distance as north-south distance. Nevertheless, it is posted north/south.

This is even more bizarre, given that WV-4 was once the longest route in the state, running from the Kentucky border over the Big Sandy River at Kenova -to- the Virginia state line east of Capon Bridge.  Almost entirely multiplexed with other routes, the one non-multiplexed section from Clendenin -to- Sutton was posted as north-south back then, as well.  But the remainder of the route was certainly west -to- east.  I'm old enough to remember the multiplexed sections, almost always posted with the WV-4 shield beneath the other shield (with no directional banner for WV-4).  So one could make the argument that the portions of WV-4 that were multiplexed with US-60 and US-50 were "posted as east/west".

However, there is historical rationale for WV-4 being posted as north/south.  Before I-79 arrived, Elk River Road was the main route between Charleston -and- West Virginia University.  Some 8 hours before Saturday football gametimes, Elk River Road was quite congested heading [north] to Morgantown.  Another oddity is that WV-4 between Sutton -and- Ireland is the only remaining section of the original WV-4.  The original WV-4 was definitely a north/south route.

I should mention that the east-west WV-4 was just a hair over 350 miles long.   Its southernmost point on US-60 near South Charleston is only about 60 miles south of its northernmost point near New Creek (or about 165 miles of diagonal heading WSW -to- ENE).

[ilpt4u]

US 150 between Vincennes, IN and Danville, IL runs predominantly N-S, partially multiplexed with US 41 in Indiana and IL 1 in Illinois.

From Danville to the Quad Cities, 150 parallels I-74, so 150 also has the N-S stretch between Galesburg and its Northern/Western terminus at US 6 in Moline near the Quad Cities Airport

[Rothman]

The 69-mile stretch of WI-13 between Parkland (US-2/US-53) and Red Cliff along the Lake Superior shore is signed north-south, even though "sbd" travels east-northeast and vice-versa.

Uh huh...

https://www.aaroads.com/forum/index.php?topic=34499.msg2929281#msg2929281

[SkyPesos]

Not the longest, but adding I-70 Breezewood to Hancock (27 miles) to the mix here.

Surprised to not see I-26 mentioned yet. Southeast of the I-385 split is debatable as it's at a diagonal, but from the 385 split to US 74 at Columbus, NC, then Flat Rock northward (including the future extension) is much more north-south than east-west.

Part of I-24's diagonal also looks more north-south, especially in IL.

[Bickendan]

I-5 has three distinct east-west segments, not counting its long diagonal run from Mexico to I-580:
CA 99 north of Sacramento to Woodland, Central Point to Grants Pass, and the overall Olympia to Tacoma segment.
Sacramento to Woodland is 13 miles, not counting the diagonal that takes it to CA 20.
Central Point to Grants Pass is 23 miles, not counting the diagonal segment through the Rogue Valley starting in Ashland, and the portion just north of Grants Pass.
Olympia to Tacoma bends it a little by including the Fort Lewis to WA 16, but comes in around 32-33 miles adjusting for the five mile northward segment.

A quick estimate of the Mexico-I-580 diagonal puts it at a 200 mile east-west change.

[michravera]

I-5 has three distinct east-west segments, not counting its long diagonal run from Mexico to I-580:
CA 99 north of Sacramento to Woodland, Central Point to Grants Pass, and the overall Olympia to Tacoma segment.
Sacramento to Woodland is 13 miles, not counting the diagonal that takes it to CA 20.
Central Point to Grants Pass is 23 miles, not counting the diagonal segment through the Rogue Valley starting in Ashland, and the portion just north of Grants Pass.
Olympia to Tacoma bends it a little by including the Fort Lewis to WA 16, but comes in around 32-33 miles adjusting for the five mile northward segment.

A quick estimate of the Mexico-I-580 diagonal puts it at a 200 mile east-west change.

People often don't comprehend that LA is only a few miles further south than east of San Francisco.
US-101 has some sections in Ventura and Western LA counties where the southbound route runs so far toward the west that a couple of turns to dodge hills actually run to the north.

[Bickendan]

I-5 has three distinct east-west segments, not counting its long diagonal run from Mexico to I-580:
CA 99 north of Sacramento to Woodland, Central Point to Grants Pass, and the overall Olympia to Tacoma segment.
Sacramento to Woodland is 13 miles, not counting the diagonal that takes it to CA 20.
Central Point to Grants Pass is 23 miles, not counting the diagonal segment through the Rogue Valley starting in Ashland, and the portion just north of Grants Pass.
Olympia to Tacoma bends it a little by including the Fort Lewis to WA 16, but comes in around 32-33 miles adjusting for the five mile northward segment.

A quick estimate of the Mexico-I-580 diagonal puts it at a 200 mile east-west change.

People often don't comprehend that LA is only a few miles further south than west of San Francisco.
US-101 has some sections in Ventura and Western LA counties where the southbound route runs so far toward the west that a couple of turns to dodge hills actually run to the north.

Another point of consideration is that the total east-west displacement I-5 undergoes is roughly 325 miles between its easternmost point (the US/Mexico border) and its westernmost points just north of Merlin, north of Grants Pass, and that San Ysidro and Tijuana are roughly the same longitudinal line as the Oregon/Idaho border.

[WestDakota]

South Carolina has plenty of them.  US 52 is signed E- W but is N- S in the state.

US 52 is strange during its whole route, going from South Carolina to North Dakota.  But the eastern end must be close to a winner for this topic, almost 400 miles from its start in South Carolina through North Carolina and Virgina as a north/south road before finally going primarily west.

[CNGL-Leudimin]

Here in Europe we don't do directional banners, so determining such sections may be hard. For example I always have a hard time thinking which directions N-234 (running from Sagunto to near Burgos) would have, I consider it to be a North-South route, however in the province of Soria, and especially "North" (West) of the namesake city, it runs East-West.

[Hunty2022]

US 29 in VA is mostly east-west from the moment it enters the state from DC to US 17 near Warrenton, but is signed north-south.

VA-3 from VA-200 in Kilmarnock to its eastern end at US 17/VA-14 in Gloucester is mostly north-south, but is signed east-west.
VA-6 from its western end at US 250 to US 29 is mostly north-south but is signed east-west.
VA-20 from its northern end at VA-3 to US 15 in Orange travels east-west, despite being signed north-south.
Almost all of VA-208 goes north-south, but is still somehow signed east-west.

[Roadwarriors79]

I-10 between Phoenix and Tucson is more of a north-south road, or at least diagonal.

[bootmii]

US 101 has two key east-west segments:

1. in California, from Route 134/Route 170 in North Hollywood to Gaviota.  122 miles

2. in Washington, from Beaver to Blyn.  81 miles

You're forgetting where it cuts across to the coast in Humboldt.

[TheStranger]

US 101 has two key east-west segments:

1. in California, from Route 134/Route 170 in North Hollywood to Gaviota.  122 miles

2. in Washington, from Beaver to Blyn.  81 miles

You're forgetting where it cuts across to the coast in Humboldt.

Drove up 101 from SF to the Oregon border a few weeks ago as part of a roadtrip, and I can't think of a significant east-west portion of 101 in Humboldt County outside of within Eureka itself.

Garberville to Loleta primarily heads northwest/southeast, Arcata to Klamath is mostly north-south.

The next significant 101 east-west portion I can see on the map is from Amanda Park, WA to Queets, WA, about 25 miles.

[Hunty2022]

South Carolina has plenty of them.  US 52 is signed E- W but is N- S in the state.

US 52 is strange during its whole route, going from South Carolina to North Dakota.  But the eastern end must be close to a winner for this topic, almost 400 miles from its start in South Carolina through North Carolina and Virgina as a north/south road before finally going primarily west.

I was shocked to find out US 52 made it to North Dakota. This whole time I thought it ended at the Great Lakes.

[noelbotevera]

Although I-81 generally travels in a northeasterly direction, in Pennsylvania there is a lot more "east" than "north". Part of this is because the road travels in a valley, part also being that PennDOT decided not to have I-81 parallel US 11 in the middle of the state. Therefore, from Harrisburg to Hazleton (~78 miles), I-81 travels eastward (spawning the east-west I-78) until it can head north to the NY state line. Conveniently, I-81 comes back to US 11 about 20 miles north in Wilkes-Barre.

Some other quick examples:

-PA 120 is signed east-west but basically follows a river. Therefore, about 47 miles (or 45% of its length) of PA 120 travels north-south.

-PA 44 is signed north-south but follows a NW-SE trajectory south of US 6. After a short concurrency on US 220, it travels east-west (and intersects the N-S US 15 near Lewisburg) to end at I-80. This stretch is about 47 miles, though you could argue that the last 7 miles are north-south.

-US 209 is signed north-south but travels due east-west until Pottsville. Then it curves northeast towards Tamaqua, but it's not until Stroudsburg that US 209 actually travels north-south. Being loose, about 110 miles of US 209 are east-west (Millersburg - Stroudsburg); but if we stay strict (counting Millersburg - Pottsville) then US 209 travels east-west for 44 miles.

[dantheman]

Since I haven't seen it mentioned yet, I'm going to say MA 28 on Cape Cod which is signed north south from Falmouth to Chatham, but runs entirely east west. Even the segment from Middleborough to the Bourne Bridge runs more east west. MA 28 just has a really bizarre path in general.

Yup. The tail end of MA 28 even ends up 180 degrees (or worse) off of its nominal direction, with the stretch from Chatham to Orleans signed "south" but running mostly north for about 9 miles.

Oddly, the same number in New York has the same issue. NY 28 is signed north/south for its entire length but actually does a big "C" shape. There's about 86 miles that are more-or-less east/west from a curve near McKeever, NY to its northern end near Warrensburg, and 84 miles from North Franklin at NY 357 to the southern end in Kingston.

The NY 28 example is clearly signed north/south throughout, which is disorienting when you leave Warrensburg going compass NW but with the highway signed "South." I think that in the 90s and early 2000s, this wasn't the case - I can't remember if that stretch was signed east/west or just didn't have directions signed at all. I can't find any evidence of this online.