RabbitEars.info - US TV Station and Market Database

It is the American dial written down as rows. RabbitEars.info - US TV Station and Market Database is an independent, single-editor reference to US over-the-air television, operating since 2008 as part of the Worldwide TV-FM DX Association, and its core is deceptively simple: every US designated market area listed with the stations and subchannels licensed within it. The market list runs 604 rows across roughly 210 DMAs, each carrying its rank, market name and total population - New York first at 19,302,889, Los Angeles second at 17,206,901. Per-market city tables take resolution below the DMA border with Census counts for constituent cities; station records then pair the viewer-facing display channel with the physical RF channel, the community of license and transmitter coordinates.

Around that core sit specialist lists almost nobody else keeps in one place: TSID records with raw TSID and TSR date views, an ATSC 3.0 NextGen TV deployment list dated by RENDER milestone, largest stations by maxcontour, deleted stations, channel changes and expirations, VHF Nightmares, channel-sharing arrangements, cross-converted networks, MPEG-4 stations, SD 5.1 operations, datacasts, ancillary service filings and a Canadian roster. The reference also computes its own READS rankings from Longley-Rice coverage predictions and Census population rather than Nielsen figures, publishes a network grid of what reaches which market over the air, and runs location-based signal search plus a live bandscan network streaming from tuners scanning nationwide.

Inside the Datadory catalog - 1,744 datasets averaging 7.81 on the quality rubric - this record scores 8/10, carried by verified field documentation and four nested levels of resolution. Get a sample of this dataset and we return rows shaped exactly like the ones below, cut to your markets.

Three cuts of the New York story, exactly as they arrive:

# market list - 604 rows, every US DMA, ranked by population
rank : 1     market : New York City, New York       population : 19,302,889
rank : 2     market : Los Angeles, California       population : 17,206,901

# cities within a market - New York market excerpt
city        : Bridgeport, CT                        population : 144,229

# stations in the New York market - viewer channel vs transmitting channel
call_sign : WCBS-TV   community : NEW YORK, NY   display_ch : 2    rf_ch : 36
call_sign : WABC-TV   community : NEW YORK, NY   display_ch : 7    rf_ch : 7
call_sign : WPXN-TV   community : NEW YORK, NY   display_ch : 31   rf_ch : 34

# specialist lists ride alongside the core:
# tsids | atsc 3.0 render dates | deleted stations | channel shares |

Read the anatomy rather than the numbers. First, the hierarchy is real: market rows, city rows and station rows share keys, so a DMA rollup, a city-weighted view and a station lookup come from one schema instead of three reconciliations. Second, the display-versus-RF split is the reason this data exists - WCBS-TV dials as channel 2 and transmits on RF 36, WABC-TV happens to match at 7 and 7, and WPXN-TV sits at 31 displaying what transmits on RF 34. Third, the specialist lists travel with the core rather than behind it, which is what turns a static station table into a dial that can be watched over time.

Eight verified fields carry every delivery, splitting into three jobs. The market frame: Rank, Market Name and Total Population - the DMA ladder every US television question ultimately climbs. Intra-market geography: City and Population, the Census counts that let a market argument get argued city by city. Station identity: Call sign, Display Channel and Digital Channel, with the display-versus-RF distinction kept explicitly rather than collapsed - the single most consequential column pair in post-repack broadcast work.

Beyond the spine, the reference's specialist lists hold real structure but uneven shapes, so they fold under additional fields on request: TSID and TSR date views, ATSC 3.0 RENDER dates, READS ranking values, maxcontour and channel-sharing references, deleted-station and expiration records, and the network-grid, signal-prediction and bandscan surfaces. Each gets mapped and populated when your sample is cut rather than promised blind here.

• Geography - the United States end to end: roughly 210 designated market areas across 604 market-list rows, resolved downward to constituent cities and licensed stations, with Canadian broadcasters carried on a separate ISED-sourced list. DMA borders are the geography US television still plans by.
• Temporal - a living reference corpus kept since 2008 and current to August 2026, with historic 1952 allotment records retained beside today's dial. Longitudinal depth accrues across repeated deliveries at whatever cadence you set - the cadence conversation is yours, not a fixed schedule.
• Granularity - four nested levels: market, city, station/facility and subchannel, with transmitter coordinates carried to six decimal places. Signal-radius and coverage math starts from real transmit points rather than market centroids.

The honest caveat: population figures are published estimates, suitable as weighting inputs rather than census-grade truths, and the strongest fit is 'what exists on the dial right now' rather than decades of history.

API, files, or your warehouse. Daily, weekly, or hourly.

You pick the channel and the cadence; the field dictionary above travels unchanged through all three. Files suit overnight loads into BI and warehouse tables, structured responses suit availability dashboards and station lookups that need to breathe with the dial, and warehouse delivery puts the station table one join away from your own market or customer data. Rows arrive flattened - market, city and station layers joined on their keys, display and RF channels typed as integers, nothing left inside markup to strip. Cadence changes are a settings conversation, not a re-integration project, and a sample cut to your markets comes first either way.

• NextGen TV transition tracking - follow ATSC 3.0 deployments station by station against their RENDER dates instead of press-release anecdotes.
• DMA footprint and reach sizing - weight markets by published populations and resolve them to constituent cities for defensible reach math.
• Antenna-decision and OTA availability tooling - pair display-versus-RF records with transmitter coordinates and signal predictions for cord-cutting calculators and installer quoting.
• Tuner and guide-data provisioning - key equipment configuration on TSID records so receivers match virtual channels to their carriers correctly the first time.
• Dial-change monitoring - watch channel changes, expirations, deletions and channel-sharing arrangements land as they happen.
• Network availability mapping - read the grid of which networks reach which markets over the air as the base layer for distribution-gap analysis.

Developers and data-product builders land at relevance 3 - antenna apps, availability tools and station lookups are exactly the products this data powers, and typed rows with a documented display-versus-RF distinction save weeks of parsing. Competitive-intel product teams and market researchers and consultants sit at relevance 2: station churn and subchannel launches read as competitive signals, while DMA populations and city counts anchor territory and reach studies. Journalists, academics and students also ride at relevance 2, citing a named reference behind every market rank and deployment date. Data scientists, sales and growth teams and investors and quants hold relevance 1 - propagation features, territory grounding and broadcast-group diligence respectively.

Persona-by-persona workflows expand on the developers and builders, competitive intel and data science pages.

It answers what is on the dial; the neighbors answer adjacent questions. The FCC TV Query System is the regulatory layer - roughly 30,000-plus authorised facility records with effective radiated power, antenna heights and licensee entities, organized by filing rather than by viewer experience. This dataset organizes the same dial by market and by what viewers actually tune, adds subchannel detail and the ATSC 3.0 timeline the registry does not frame. On the demand side, Barb's UK products measure what was watched, not what transmits; the Internet Archive TV News Archive holds what was said, at caption-searchable grain. The head-to-head with the measurement side is worked through in BARB Viewing Data Hub vs RabbitEars.info.

Four things, all knowable upfront. First, populations are estimates - useful denominators for weighting and reach work, not census-grade counts, so treat derived per-capita measures as directional. Second, this is a curated reference, not a registry of record: breadth comes from one editor's discipline, and regulatory-authority questions belong to the FCC pair, where filings are the ground truth. Third, specialist lists vary in shape - TSID views, RENDER dates and bandscan reads each carry their own grain, so scope them explicitly at sample time and the delivered schema extends to match. Fourth, history has a boundary: deep coverage of the current dial plus 1952 allotments, with longitudinal series accruing from repeated deliveries - start the cadence before the quarter you need the trend.

Because a reference this good deserves better than being retyped by hand, one market at a time. Datadory resolves the whole thing into typed rows: the market-city-station-subchannel hierarchy joined on its keys, display and RF channels distinguished by construction rather than convention, TSIDs and deployment dates attached to the stations they belong to, and the specialist lists scoped to whatever your project actually touches. Start with a sample cut to your markets and stations, then browse the rest of the slice on the broadcasting data hub or the best broadcasting datasets ranking.