How are zip codes in the US assigned?

If I get your meaning of what you are trying to do I don't think it's possible.

The first 2 digits of the zip code correlate to the region but the last 3 are just assigned sequentially as postal regions are divided up.

If my current zip code is 99226 and it gets split one part would keep 99226 and the other could become any unused number between 99001 and 99999. The distance isn't relative.

Maybe there is something here that would be helpful???

http://www.usps.com/ncsc/addressinfo/addressinfomenu.htm

Grr, you both understand my idea and confirm my fears.

I don't think my client can afford a zip database liscense and the coding that it would take to use so they'll be stuck with searching for exact or partial zipcodes (with no proximity calculations).

Could you incorporate something like this?

http://www.cryptnet.net/fsp/zipdy/

I could, but not for their price. It's have to be converted for MySQL and the coding would be extensive.

Incorporating that code (once the DB data is converted) would be easy, but to make the search return their listings based on proximity is more coding than they want to pay for.

And that database probably isn't as accurate as most commercially maintained ones.

I'll throw it out to them but that'd increase the price a lot. I think they'll just go for basic zip search with no proximity.

I just checked their demo of the DB, it didn't have my zipcode...

I'll just tell them that if they want it they'll have to pay. If not I'm ready to do what we already agreed to.

There's a little more to it -- the third digit depends on whether you're in a city center or not... 1 is a major city, 0 is a small town or unincorporated area.

I don't think the last two are totally random either. I think they were originally assigned to a certain geographic region (i.e., the area covered by a single post office). Do they split zip codes like they do area codes, when more people move into an area, or is that handled by the additional four digits? I know they are mostly for businesses, but homeowners are assigned them too. They (the four extra numbers) are quite specific for businesses, but for personal use I think they refer to neighborhoods...

I used to work for a company that widely distributed self-addressed envelopes with the zip printed as a bar code at the bottom of the envelope. People would cover the address with a label and write on their own address (not trying to steal postage, just to use the envelope for personal mail). If they didn't cover the barcoded zip, the envelope came to our business, no matter what address was handwritten on the envelope...

Maybe some ideas can be garnered from this link.
http://www.fammed.washington.edu/wwamirhrc/rucas/methods.html

Ok, I was just being brief earlier but the 1st digit indicates the "major region" within the US. (All of New England is "O" and the Northwest all start with "9", "7" is TX, OK area, etc..). The 2nd and 3rd digits are assigned to postal sorting centers in major city areas within each region. The last two digits are sequentially assigned to each of the individual post offices serviced by the sorting center.



They are by Post Office but the sequence of which Post Office gets which number is pretty much random.



If an area gets big enough they do split them up. Usually the Post Office will open a branch office to handle the mail load and eventually they'll assign a unique zip code to that office if the population keeps growing. The +4 addressing is, as you mentioned, either by business or by neighborhood. They can sort mail right down to the individual mail carriers bag with the +4 system in some towns.

http://freshmeat.net/projects/zipdy/

http://freshmeat.net/projects/zipdy/

oops

I'd seen that project. None of the zipcodes I entered worked correctly.

I'll give it a look in the future to see if it can be improved though.

The postal zones were invented circa World War I. In 1959, the Zone Improvement Plan (Zip Code) was introduced. In my experience of working with data bases for mailing lists, the numbers are assigned to an alphabetical listing of town names within the previously existent zones. So, for example, 43- is central Ohio, 432 is Columbus, 433 is southeast central Ohio, etc. Within 430, 43017 is Dublin, 43026 is Hilliard, etc. The gentleman who proposed the zone improvement plan (whose name escapes me), simply proposed adding digits to the existing zones assigned to alphabetic listing which would be easy for the punch-card computers of the day to handle, and would allow the readers of the day to read the primitive bar codes which you will still occasionally see on you letters. When you see that, it simply means the letter addressed to you was on the top of the bundle, and the bar code was printed on the envelope to facilitate machine sorting at each station through which the bundle passes.

As for your problem, Crave, i'm sorry to say i don't think there's any cheap way around it. When i worked with mailing lists, it was pay through the nose, or create your own zip data base.

The only thing I can offer is that they increase east to west, and south to north, seemingly backwards. If I had had anything to do with it, 00000 would be The Bering straights.