Data report
US phone line types, counted. Every assigned block in the US and Canadian number plan.
Of the 1,539,378 thousand-number blocks assigned across the US and Canada, 40.5% belong to wireless carriers, 31.3% to landline carriers and 28.2% to VoIP providers. Number pooling splits one exchange between carriers, so 47.5% of exchanges hold more than one kind of line - which is why a phone's type cannot be read from its area code alone.
The number plan. Every assigned block, by the kind of carrier holding it.
A block is a thousand consecutive numbers, and this is the shape of the number plan rather than of the phones in use - nothing here says how many people carry a mobile. It says what a carrier filed for. A number ported since it was assigned keeps its block, so only a live portability lookup, which asks where the number sits today, can show the move.
One exchange, two kinds of line. Which is why an area code and prefix cannot answer the question.
An exchange - the area code plus the next three digits, ten thousand numbers - used to belong to one carrier. Since number pooling it can be split ten ways, a thousand at a time, and the parts can be different kinds of line: a wireless block beside a landline block inside the same exchange. 142,003 of the 231,623 exchanges (61.3%) are split that way, and 109,942 of them (77.4% of the split ones, 47.5% of all exchanges) hold more than one kind. 85,828 hold at least one wireless block - those are the exchanges where a guess costs you a mobile.
Guess at the exchange and 25% of blocks are wrong. Measured, not asserted.
Give every block the most common line type in its own exchange - the best an answer that stops at the prefix can do - and 327,459 of 1,307,755 blocks get the wrong kind of line. One in four. The SQL is a window function over the same table and it is printed in the file this page renders from, so the figure can be reproduced rather than taken on trust.
Blocks, not phone numbers. What share of real numbers that works out to depends on how numbers are distributed inside each block, which the filings do not say - so neither do we. It is the reason Leadity resolves a phone against its own block and answers "unknown" when the blocks disagree and the number's own is missing, rather than taking the exchange's word for it.
By state and province. Regions holding 5,000 blocks or more.
Method. After the argument, where it belongs.
The block assignments are published by localcallingguide from NANPA's filings and are free to anyone. This platform holds a copy - 1,539,378 rows, loaded 2026-09-20 - because classifying a scraped phone against a local table is what keeps a scrape from depending on somebody else's rate limit. Each row is one area code, one exchange, one thousand-block digit, and the carrier that holds it.
A block becomes a line type by its carrier type: W is wireless, I is a landline incumbent, and a competitive carrier is decided from its name against a keyword list - the same rule the product uses on every number it delivers, so this report describes the thing we actually sell. Figures read 2026-09-22. Every query sits beside its figure in the file this page renders from.
Reuse it. Quote any figure or table here with a link back to this page. If you want the aggregates as a file, ask and we will send them.
Questions.
No. An area code says where a number was assigned, not what kind of line it is. Even the full exchange - the area code plus the next three digits - is not enough: 47.5% of exchanges hold blocks belonging to carriers of different kinds, because numbers are allocated a thousand at a time and the parts can be split between a wireless carrier and a landline one.
Numbers are allocated to carriers in blocks of a thousand: the area code, the three-digit exchange, and the first digit of the last four. 415-555-3000 through 415-555-3999 are one block. Before pooling an entire exchange of ten thousand numbers went to a single carrier, which is why the older shortcut - look up the exchange - used to work.
It is as accurate as the carrier filings it comes from, and it says who the number was ASSIGNED to. Number portability breaks that: a number carried from a landline to a mobile keeps its block. This report counts what the blocks say, over 1,539,378 of them, and does not claim to know where a phone is now.
The block assignments are published by localcallingguide from NANPA's own filings and are free to anyone. What is ours is the counting, and the rule that turns a block into a line type - a carrier type of W is wireless and I is landline, and a competitive carrier is read from its name against a published keyword list. The SQL for every figure is in the file this page renders from.
Because a list sold as mobile numbers is only as good as the lookup behind it. Leadity classifies every number it scrapes against this table at the block level, and delivers a lead only when the block says mobile - which is what a carrier-verified mobile from $0.016 means.
Related: the Google Maps scraper and the rest of the guides.
Leads with the line type already checked.
Pick a niche and a US or Canadian area, keep only the mobiles, and export the list - from $0.016 a carrier-verified mobile.