I've read: Why have different flights with the same flight number?

and I guess I understand this can make sense. But - is it allowed for a domestic US airline to schedule two flights with the same number, on the same day, with intersecting routes? Specifically, which have legs scheduled to fly out from the same airport on the same day?

This would be potentially quite confusing for passengers, even if the departure times are spaced many hours apart.

Motivation: I'm seeing what seem to be records of this occurring a few times in a USDT flight data table. Concrete example (edit: this is a significant example, but the question is about the general case) - some US airways flight legs from 2006:

uniquecarrier flightdate flightnum origin dest crsdeptime crsarrtime tailnum deptime arrtime cancelled
US 2006-01-25 634 LAS PHL 645 1416 N632AU 642 1407 false
US 2006-01-25 634 LAS PIT 2257 607 N808AW 2252 556 false
US 2006-01-26 634 LAS PHL 645 1416 N621AU 650 1428 false
US 2006-01-26 634 LAS PIT 2257 607 N833AW 2303 603 false
US 2006-01-27 634 LAS PHL 645 1416 N633AU 643 1404 false
US 2006-01-27 634 LAS PIT 2257 607 N832AW 2249 545 false
US 2006-01-28 634 LAS PHL 645 1416 N629AU 645 1404 false
US 2006-01-28 634 LAS PIT 2257 607 N806AW 2256 544 false
US 2006-01-29 634 LAS PHL 645 1416 N602AU 640 1409 false
US 2006-01-29 634 LAS PIT 2257 607 N827AW 2256 602 false
US 2006-01-30 634 LAS PHL 645 1416 N632AU 641 1422 false
US 2006-01-30 634 LAS PIT 2257 607 N801AW 2250 536 false
US 2006-01-31 634 LAS PHL 645 1416 null null null true
US 2006-02-01 634 LAS PHL 645 1416 N628AU 652 1420 false
US 2006-02-01 634 LAS PIT 2257 607 N810AW 2300 549 false
  • $\begingroup$ Related, possibly dupe: aviation.stackexchange.com/questions/44416/… $\endgroup$ – Ron Beyer Feb 1 at 22:59
  • $\begingroup$ @RonBeyer: I linked to that question on the first line of my question... $\endgroup$ – einpoklum Feb 1 at 23:04
  • $\begingroup$ It's quite possible you're seeing an error in the data. Can you post some here? $\endgroup$ – CatchAsCatchCan Feb 1 at 23:17
  • $\begingroup$ @CatchAsCatchCan: Yes, that's exactly what I'm wondering myself. Posted an example, or rather a few examples. $\endgroup$ – einpoklum Feb 1 at 23:23
  • 1
    $\begingroup$ Your bounty says, "help me figure out what's going on with this dataset" and then you edit the question to de-emphasize the data as merely the genesis of the question about the general case... it's your question, but this looks like you're shifting it back & forth a bit. $\endgroup$ – Ralph J Mar 14 at 20:57

Short Answer:

Flight 634 LAS-PHL at 0645 was most likely "US Air 634" on those dates, while flight 634 LAS-PIT at 2257 was most likely "America West 634" on those dates. Even if both were to have flown with the ATC callsign "Cactus 634," that wouldn't be a problem for ATC since they were separated in time by several hours. (No more an issue than N12345 flying KABC to KABC more than once in a day.) But most likely, they flew with separate callsigns too, as explained below.

Why is this the likely answer?

The LAS-PHL flights in the OP were flown by aircraft whose registrations end in "AU" such as N628AU for example. The airfleets website shows that this aircraft was a US Air (later, US Airways) jet until it was re-registered in December 2006 (the dates on that site are DD-MM-YYYY format). Generalizing from that example, these jets were legacy (pre-merger) US Airways jets.

The LAS-PIT flights in the OP were flown by aircraft with registrations ending in "AW" such as N810AW. Airfleets shows that this was an America West jet until it became a US Airways jet in October 2006. Again generalizing from this example, the "AW" jets were "American West" aircraft at the time of the OP's data.

So at the time, the one flight was being flown by jets in US Airway livery, and the other by jets in America West livery. And to the passengers, they almost certainly were sold under the respective brands, so one could buy a ticket for "US Air flight 634" from LAS to PHL in the morning, or a ticket on "America West flight 634" from LAS to PIT in the evening. You'd have to use two separate websites to do so, but there would be no confusion.

Okay, so the remaining question:

Why do they both have the "US" (presumably meaning US Air) uniquecarrier ID in the table?

As described in this paper from Embry Riddle Aeronautical University,

America West Airlines acquired the bankrupt US Airways on September 27, 2005 to form the US Airways Group.

The two airlines became fully one in 2007, when a single operating certificate covered them both. During the interim,

Both of the merged airlines retained their names and continued operating until September 25, 2007, when America West's FAA certificate was merged into US Airways. (source)

This interim period is when the flights shown in the OP occurred. At this point, the airlines were in the process of merging their operations, although their website weren't merged until May 2006. Thus, in the January/February timeframe posted, the tickets for each would be sold separately.

The wikipedia article notes that the callsigns remained separate until 2008, as "US Air" and "Cactus"; after October 2008, the entire operation used "Cactus".

While it's hard to retrace each step along the way from two separate carriers to one unified carrier, it's clear that various steps are involved, so it seems probable that whatever data source gives the uniquecarrier identifier in the table, either switched over separate identifiers before the dates of these flights, or else it was retroactively changed over to make them all "US" even though at the time some would have been "AW".

In the general case...

So, is it possible, to have two flights with the same number at the same airport at the same day? Yes, although the data shown above probably doesn't indicate that LAS saw a "US Air 634" twice a day during the timeframe shown. But as long as the first flight and the second don't overlap in time, there is no ATC constraint that prohibits is. (Military aircraft do this all the time... Cool06 can fly from KDLF to KDLF on a training mission multiple times during the day -- Laughlin AFB in this case.) It's possible, but sufficiently unwise because of the potential to create confusion, that it probably happens for an airline somewhere between very rarely, and never.

As a matter of avoiding confusion, an airline would probably work to avoid scheduling their flight 123 more than once in a day:

"I'm here for my flight 123."
"Sorry, but YOUR flight 123 isn't for another 10 hours; this is the other flight 123 today."

But, if flight 123 is scheduled to depart at 2300 and one day it runs over an hour late, the next calendar day will have two "Flight 123" operations, and nothing will break.

All that to say, what the data in the OP most probably shows, is an artifact of how US Air operated during the transition from two airlines into one, and may not really be particularly indicative of anything more widespread, nor generalizable, than that.


From an air traffic control perspective, how the flights are referenced when marketing to passengers is not important. Instead the flight, or more correctly the aircraft, callsign is what matters.

Sometimes a flight will have an ATC callsign that is not the same as the flight number. For example Finnair operates a flight from Helsinki to Paris that is advertised as "Finnair Flight 1577" and likely displayed to customers as "AY1577" (AY being the IATA airline code for Finnair). But on ATC displays the flight shows as "FIN3EF" (FIN being the ICAO airline code for Finnair) and uses the callsign "Finnair Three-Echo-Foxtrot." This system seems to be used with more regularity among European airlines than American.

In the US, the callsign is almost always the flight number, but it doesn't have to be. One example I've noticed at my airport: There is a late-night arrival XYZ5163 from one airport, and an early-morning arrival XYZ5342 from the same airport. If the late-night flight is delayed long enough that they decide to run it the next morning instead, it will often have a callsign like XYZ534A, despite (I assume) having last night's flight number. ATC is also supposed to be alert for similar-sounding callsigns; if a controller notices confusing similar callsigns they're supposed to let management know (7110.65 2–4–15) and management will reach out to the airlines and see if there's a way to change one of the callsigns (7210.3 2–1–14).

So: as long as there is never more than one flight with the same callsign in the air at any one time, ATC doesn't care what the flight numbers are. That's as far as I know for the "faa-regulations" tag on your question. As for the marketing/passenger side, I think the other answers pretty well explain that USA and AWE were operated as distinct airlines until 2007. The Wikipedia article for AWE even says:

Beginning in January 2006, all America West flights were branded as US Airways, along with most signage at airports and other printed material, though many flights were described as "operated by America West." Apart from two heritage aircraft, the only remaining America West branding on aircraft were found on some seat covers and bulkheads. The merged airline used America West's "CACTUS" callsign and ICAO code "AWE", but retained the US Airways name.

So I think it's likely that whoever made your dataset assigned the "US" IATA code to both sides of the operation, or perhaps the operation itself used the "US" code for both, while ATC continued to see separate "AWE" and "USA" ICAO codes.

  • $\begingroup$ But what about the general case? You addressed the example. Is it never the case that the same flight number, of the same carrier (presented the same way to the passengers) $\endgroup$ – einpoklum Mar 14 at 9:26
  • $\begingroup$ @einpoklum I can't speak well to that. Certainly what Ralph said is accurate, sometimes a flight scheduled to depart before midnight is delayed until after midnight—then you have two flights with the same number and destination that departed on the same calendar day. But I would guess that airlines try not to advertise them as departing on the same calendar day. I'm not actually certain how the flight number was described to customers in my "XYZ534A" example. $\endgroup$ – randomhead Mar 14 at 16:07
  • $\begingroup$ The crosses-midnight situation is not of interest to me, since I care about the scheduling (as you can tell from my example, my data is based on scheduled dates, not actual flown dates). And I also don't care about arrivals-cum-departures - just pairs of scheduled departures. $\endgroup$ – einpoklum Mar 14 at 16:26

Yes it is possible to have flight with same number departing same airport at the same day when they have different flight number. By flight number I mean airline code combine with number. For example London Heathrow may have Imaginary Flight BA1, SQ1, CX1 and EK1 on the same day but they are different flight clearly mark by different airline code.

From your example, the pattern is clear. The flight to PHL operates by aircraft ending in AU while flight to PIT end with aircraft ending in AW. Brief google around reveals that aircraft ending with AU belongs to US Airline and AW belongs to American Airline. So it is just a case of different airline with the same number on the same route but differ by airline code. You need to pull airline name to see this pattern.

  • $\begingroup$ small issue, the tail numbers all correspond to US Airways at the time of the data, e.g.: planespotters.net/airframe/… $\endgroup$ – ymb1 Mar 13 at 22:43
  • $\begingroup$ All these flights are listed under 'US Airways' in my dataset! Updating the table to better clarify this. Are you sure about the suffixes and airline association? $\endgroup$ – einpoklum Mar 13 at 23:41
  • $\begingroup$ For a while, "US Air" included both "US Air East" - traditional US Airways, and " US Air West" - America West Airlines (which had actually bought US Air). This was long before the combined entity was bought by American. One could hypothesize that one flight was sold as "US Air 123" while the other was sold as "America West 123". The convention at the time may have been that both operations got the same "unique carrier" ID (and perhaps flew as "Cactus123" - though widely separated in time), even though the public saw 2 separate brands. $\endgroup$ – Ralph J Mar 13 at 23:55
  • $\begingroup$ A google search on N628AU shows that 757 in old-style US Air livery, and N810AW as being an America West bird until 2007. I suspect that the data in the OP is an artifact of how USAir operated during the transition from 2 airlines into 1, and may not be particularly interesting, nor indicative of more widespread nor generalizable trends, beyond that. $\endgroup$ – Ralph J Mar 14 at 0:02
  • $\begingroup$ @RalphJ: So, the question remains - what about the general case? $\endgroup$ – einpoklum Mar 14 at 9:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.