How are 3 digit interstate #'s determined?

[bluecountry]

Obviously even #'s are for loops/bypass that reconnect at a second point with the parent interstate while roads which do not reconnect to the parent and spur are odd numbers.
But:

1.  How do they determine if a highway is 284, 484, or 684 all things being equal (such as no other route with that designation)?
2.  If a highway starts at one interstate (like 95) and connects to another (like 64 hypothetically), is that considered a loop since it does connect to another interstate or a spur?

[DTComposer]

For question 2, there's several examples in California of even 3dis not terminating at its parent:

I-280 - never touches its parent (although it was planned to) - termini are I-680 and King Street
I-680 - termini are I-80 and I-280
I-880 - termini are I-80 and I-280

I-205 - termini are I-5 and I-580
I-605 - crosses I-5, but termini are I-210 and I-405

I-210 - termini are I-210 (in both the AAHSTO and the CA-210 versions) and I-5

Several of these (880, 205, 605) could perhaps more correctly be considered spurs. To me I-605 definitely feels like two spurs from I-5 rather than part of a bypass/loop.

[epzik8]

For the first question, I'm not sure if there's a specific method traditionally used for determining the first digit, aside from maybe how important the area served by said highway is, but that's a completely wild guess of mine.

[roadman65]

New York does it best with x-90 children minus I-990 which came later. From 1 to 8 there is an order with 1-2 starting in the west and 8 in the east.

Some could argue that 3 is east of 4 as I-490 spurs west of I-390 on their parent.  However, it's close enough with I-490 extending east of I-390.

[hotdogPi]

I created a spreadsheet regarding first digit vs. length. There is almost no correlation (and what does exist is the opposite of what you'd expect; lower numbers are just barely shorter).

[roadman65]

I know I-287 in NY and NJ was supposed to be an x-95 as it should as both I-287 terminuses are at I-95 and not its parent.  However, all the digits for an I-95 child were used ( or planned to have been) at the time of numbering.

Why I-684 in Westchester County, NY is not an odd Spur is unclear, but it connects two interstates so it doesn't violate the rules.

I-495 on Long Island should have been I-395 as it never was intended to be a loop nor will it ever be.  It doesn't terminate at interstates either end presently. Originally I-495 was to spur from I-95 at Secaucus, NJ, but cancellation of the Mid Manhattan Expressway foiled that and NJ's portion became State Route 495. Its only connection to I-95 is via I-295 in Queens.  Yes, I-278 too, but in terms of siblings I-295 makes it legit to be an x-95 designation.

[DJStephens]

Am of belief that a Long Island Sound crossing was envisioned, when the route #'s were being assigned, so that explains the 495 number on what is essentially an orphaned spur.   Would have been interesting to have seen what would have played out, in terms of where the route would have landed in Connecticut on 95, had the crossing been built.     

[freebrickproductions]

I know in Tennessee, the even-numbered I-40 children increase as you go from west to east (I-240 in Memphis, I-440 in Nashville, and I-640 in Knoxville), save for I-840 around Nashville, which was added later and likely got the number due to being available at the time. Dunno how much correlation location and first digit number have with each other in general though, I-140 and I-124 in Tennessee are both in the eastern part of the state (Knoxville and Chattanooga respectively), while I-265 used to exist in Nashville.

[jeffandnicole]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

[jlam]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

Tell that to I-820

[Quillz]

Generally speaking, I'd say it's somewhat dependent on location. The westernmost spur in the state might be 180, the easternmost might be 980. But this can fall apart unless all the envisioned spurs are built at the same time.

In some cases, numbers might try to fit into the state route network. Nevada, for example, likely gave 580 that number because "5xx" is used for state routes in urban areas. So it fit in quite well.

In other cases, it could simply be sequential. The first spur is 1xx, the second one 2xx, and there may not be any adherence to odd numbers being for spurs, even numbers being for loops, etc.

Basically, there are no real standards or rules, as none of those were truly established. Merely guidelines about how numbers should be used.

[Big John]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

x94, low by Chicago, mid in Minnesota, high by Milwaukee.

[jeffandnicole]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

Tell that to I-820

Wasn't that State Route 820, then they gave it an upgrade to I-820?  I believe that's the backstory there as to how it was given an unusually high number.

[Molandfreak]

I remember reading somewhere that 820 was used to advertise a local AM radio station.

In the case of Minnesota, the numbers are simply the lowest available numbers that aren't already in use by another state highway. I don't know why they chose to stick a 1 in front of former state highways 35 and 94 rather than, say, extending TH 200 over current TH 194 and choosing a completely different number for 135 in order to free up those numbers, but a nice benefit resulting from that is there was no duplication with 3-digit interstates other states. There would have been if 335 was built in Minneapolis and 494 was built in Chicago, though.

[GaryV]

In Michigan, they go low to high from W to E (96) and from S to N (75). Same as mile markers.

[TheStranger]

California:

The original 1956-era designations were...

I-5: I-405, I-605, I-805
I-10: I-210
I-80: I-280, I-480, I-680

Then as the network expanded:
I-5: I-205 added ca. 1958-1959 (as a result of 5 being moved away from the 99 alignment in Modesto to the West Side alignment), I-505 (formerly planned as I-5W) added in 1964, as was the original I-105; current I-105 created in 1968.  Unsigned I-305 along US 50/former I-80 created in 1982, Future I-905 designated in the early 1980s (so 705 has been skipped entirely)

I-10: Original I-110 designated in 1964 and rescinded several years later.  Current I-110 created in 1981 (number derived from 1934-1981 Route 11), current I-710 designated in 1984 (number derived from 1964-1984 Route 7), which means that 310 and 510 were skipped.

I-15: I-15E became I-215 in 1982

I-80: I-580 created in 1964 on what had been I-5W/US 50, original I-880 designated in 1964 for the North Beltline in Sacramento, then removed in 1982; I-380 designated ca. 1970, I-780 created from old I-680 alignment in 1976, I-980 designated on what had been planned as Route 24 ca. 1981; current I-880 created from a potion of Route 17 in 1984 (and then there's I-238...)

Essentially, California did number their 3dis somewhat geographically sequentially at first (note that 205 is way north of 405, which has a terminus slightly north of 605's path, and then that is pretty far north from 805), then just started filling in gaps afterwards.

Looking at the example of I-80: the original sequence was 280 as the westernmost (with 80's planned terminus at 280/1 in Golden Gate Park factored in), then 480 along the waterfront, 680 inland from San Jose to Concord to Vallejo. (The Sacramento 880 fit this sequence, and the road was planned in the late 1950s, but was not given the 3di number until the 1960s)

[BJ59]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

Tell that to I-820

Wasn't that State Route 820, then they gave it an upgrade to I-820?  I believe that's the backstory there as to how it was given an unusually high number.

I-820 was built from the old existing loops in Fort Worth: Loop 217 (Present Day I-20 near US-377) and Loop 820 (From I-35W south to SH-121). In addition, new sections of highway were built to complete the loop around Fort Worth (specifically in the Northwest quadrant)

On the topic of 3 di interstates, I think Fort Worth is a candidate for another 3di interstate loop around I-820.

[SEWIGuy]

I always figured they just started at the lowest number (1xx or 2xx) then went up as more were added.  Occasionally a number may be skipped - a planned route never occurred, or an adjoining state already had the low route number so they used the next one to avoid confusion.

Wisconsin has I-894 and I-794. No other routes were planned.

[WillWeaverRVA]

I-495 on Long Island should have been I-395 as it never was intended to be a loop nor will it ever be.  It doesn't terminate at interstates either end presently. Originally I-495 was to spur from I-95 at Secaucus, NJ, but cancellation of the Mid Manhattan Expressway foiled that and NJ's portion became State Route 495. Its only connection to I-95 is via I-295 in Queens.  Yes, I-278 too, but in terms of siblings I-295 makes it legit to be an x-95 designation.

One could say that I-495 would have worked as a loop if the Manhattan section had gotten built - if the rest of current I-495 hadn't been built, you could complete the loop via I-295.

[triplemultiplex]

There are plenty of examples of states that went with the plan of numbering 3dis west to east and south to north.  And plenty of states where they just went with whatever they felt like.

The Michigan, New York and Tennessee examples have been cited already as places where there was a "system" they were following.
It was clearly on the mind of planners when they picked 3di numbers for Milwaukee going with higher first digits since the city was so far east on I-94's path thru the state.  Ostensibly, that holds lower numbers in some kind of "reserve" for future spurs and loops further west.
The progression of 185, 385, 585 in South Carolina is another example.
This is why Baltimore has all high numbered x95's while lower number x95's are around the Maryland side of DC.

Thinking about Texas in this context and its clear similar thinking was in place.
210 left open for a future loop in El Paso; 410 San Antonio, 610 Houston.
The main Dallas loop got 635 telegraphing the intent to potentially use 235 and 435 further south some day.

This insight explains why you see seeming outlier 3di numbers where they could have gone with a lower number but didn't.  They were skipping some numbers so they could backfill later if needed and still be in sequence.
Alabama: 359 and 759 match up well to add a 559 to Birmingham.
Minnesota: Duluth gets 535 instead of a lower number because it's further north. (Tho the other factor there is explained by Molandfreak)
Georgia: 985 is a massive outlier until one considers it in the context of leaving 3, 5 and 7 available for future spurs in Atlanta.

One thing is certain, 1xx and 2xx are far more common 3di numbers than any other leading first digit indicating a default practice of starting at bottom and working your way up.  (That's how we end up with two I-291's a stone's throw from each other.)
At least that's how it seems to have gone down when the first batches of 3dis were being assigned.

In recent decades, it's been more like whatever someone feels like, or based on some arbitrary in-state "rule" that isn't actually a rule, but likely the preference of a single individual at some point that everyone went along with.
Examples: Arkansas picked all 5xx's for their long spurs.
Oklahoma and their obsession with new 3xx's for their toll roads.

But every so often we get a throw back to an era where people thought about the future. The relatively new spur to Fort Drum in New York was given 781, conveniently leaving the door open for 3 other spurs further south to be in sequence.  And now as NY plans on moving 81 onto 481 in Syracuse, roadgeeks be all like, "Make the new stubs 3dis!"
If they did, the system would suggest the north stub becomes 581 and the south stub becomes 381.

[achilles765]

What instance were Texas did probably make a mistake with the designations is with interstate 345 in Dallas.
There are at least three potential spare routes that could be designated off I 45 south of Dallas.
In a perfect world, FM, 1764 in Texas City would be an IH-145, the NASa Parkway Freeway bypass would be IH-345, the Hardy toll road could be redone as IH-245 or IH-445 with the airport connector being an IH- 545

[vdeane]

I-495 on Long Island should have been I-395 as it never was intended to be a loop nor will it ever be.  It doesn't terminate at interstates either end presently. Originally I-495 was to spur from I-95 at Secaucus, NJ, but cancellation of the Mid Manhattan Expressway foiled that and NJ's portion became State Route 495. Its only connection to I-95 is via I-295 in Queens.  Yes, I-278 too, but in terms of siblings I-295 makes it legit to be an x-95 designation.

It's worth noting that the original I-495 ended at I-295; east of there was NY 495.

[TheStranger]

A few other examples that came to mind:

Nevada: IIRC the usage of I-515 and I-580 (starting with their designations in the 1970s and eventual in-field signings in the 1990s and 2010s respectively) fits in with their state route highway numbering system for some reason.

Washington: Isn't I-705 derived from WA 7?  Otherwise, straightforward geographic sequence (I-205, I-405). 

Arizona: The Phoenix area had a proposed I-410 (briefly signed) and I-510 (which led to today's AZ 51) and proposed I-710 in Tucson, which means...lots of skipped numbers (110, 210, 310)

North Carolina: I-40's progression of 3dis was originally west to east: I-240 Asheville, I-440 Raleigh, planned I-640 (now I-540) as outer loop of Raleigh.  I-840 in Greensboro is basically infill, not unlike the I-840 around Nashville

I-85 on the other hand...I-485 is the very first 3di upon entering the state, and I-285 is a much later designation near Winston-Salem, with I-885 used for the newer Durham connector with US 70 (and 685 being proposed for US 421).  The first odd 3di on the list is...I-785.

[Rothman]

I-495 on Long Island should have been I-395 as it never was intended to be a loop nor will it ever be.  It doesn't terminate at interstates either end presently. Originally I-495 was to spur from I-95 at Secaucus, NJ, but cancellation of the Mid Manhattan Expressway foiled that and NJ's portion became State Route 495. Its only connection to I-95 is via I-295 in Queens.  Yes, I-278 too, but in terms of siblings I-295 makes it legit to be an x-95 designation.

It's worth noting that the original I-495 ended at I-295; east of there was NY 495.

Ugh...you're reminding me of the headache one has when determining which shielded Interstates in NYC are eligible.for 90% NHPP and which ones are not.

[Scott5114]

Nevada: IIRC the usage of I-515 and I-580 (starting with their designations in the 1970s and eventual in-field signings in the 1990s and 2010s respectively) fits in with their state route highway numbering system for some reason.

Urban arterial routes in Nevada start with 5 or 6 and are allocated in blocks by county. Amusingly, though, 515 is a Carson number and 580 is a Clark County number. There's obviously no other way to comply with Interstate numbering, though.

This implies that if I-11 ever gets an auxiliary route in Southern Nevada, it will be I-611, since that's a Clark County number that currently isn't in use.