Where do we get our data?

Our primary dataset is built directly from the official Centers for Medicare & Medicaid Services (CMS) CY 2026 Ambulance Fee Schedule (AFS). This federally maintained database details the exact maximum allowable amounts the government will reimburse participating providers for ground and air ambulance transports across the United States.

The Power of the CMS Baseline

Private ambulance companies and municipal fire departments legally have the freedom to set their own "Sticker Prices," which are often astronomically inflated to serve as starting points for negotiations with private insurance companies. Because these sticker prices vary wildly from county to county and provider to provider, there is no reliable national average for retail transport bills.

However, the CMS fee schedule offers a concrete, federally-audited baseline. It represents what the government has objectively calculated that a transport should cost in any given locality to fairly compensate providers for their equipment, fuel, and highly-trained medical staff, while protecting patients from surprise billing.

How CMS Factors in Location

A transport in downtown Manhattan costs differently to run than a transport in rural Montana. To account for this, the CMS dataset applies specific locality modifiers to its national base rates:

• Urban Zones: Base rates are adjusted using the specific Geographic Practice Cost Index (GPCI) for that metropolitan area, factoring in elements like local rent and wage standards.
• Rural Zones: Providers operating in rural areas receive a bonus modifier. This helps subsidize emergency networks that experience lower call volumes but must maintain 24/7 readiness over vast service areas.
• Super-Rural Zones: The CMS identifies areas in the lowest 25th percentile of population density as 'Super-Rural'. Transports originating in these remote areas receive an additional 22.6% bonus multiplier to prevent the total collapse of local emergency infrastructure.

Real Billed Charges from Medicare Claims Data

A published municipal fee schedule is the gold standard, but most ambulance services in the country never post one. To fill that gap with real numbers instead of a model, we use the CMS “Medicare Physician & Other Practitioners — by Provider and Service” Public Use File — a federal dataset built from actual Medicare claims. For every ambulance provider that bills Medicare, it reports the average amount that provider actually billed for each service (BLS, ALS, and loaded mileage), tied to the provider’s NPI and ZIP code.

Crucially, this is the provider’s billed charge — the sticker price an uninsured patient faces — not the discounted amount Medicare actually pays. In the same file the billed charge runs roughly twice the Medicare-allowed amount, which is how we confirm we’re reading the charge and not the reimbursement. We validated these figures against fee schedules we had already verified by hand: provider-for-provider, the Medicare-claims charge matched the posted rate almost exactly. This one federal dataset added real billed-charge data for thousands of providers nationwide, including many communities that publish nothing at all.

Because this figure is an average of a provider’s recent Medicare claims rather than a single posted number — and reflects the most recent full year of federal data — we treat it as a high-confidence figure that sits just below a hand-verified published schedule. On the results page it’s labeled as a charge “reported to Medicare,” with a link to the federal dataset and the provider’s NPI so anyone can look up the exact record.

From Medicare Baseline to a Real-World Estimate

The Medicare figure is what the government pays — not what a private provider bills you. When we have a city or county's actual published rate, we show that exact number with a source link and the date we captured it. When we have neither a published schedule nor that provider’s Medicare billing data for an area, we estimate the sticker price, and we want that estimate to reflect your region rather than a blunt nationwide guess.

To do that, we built a model from the hundreds of verified provider rates already in our database. For each one we measure a simple ratio: how many times higher the provider's real billed rate is than the Medicare rate for the same area. Bundle those ratios together and a clear regional pattern emerges — markups in California run far higher than in Florida, and rural districts differ from dense urban ones.

• Grouped by state and locality. We calculate a separate markup for each state, split by Urban, Rural, and Super-Rural, so an estimate for a rural ZIP is based on what rural providers in that state actually charge.
• Median, not average. We use the median markup so one unusually aggressive biller can't skew the estimate for a whole region, and we cap the multiplier to a sensible range to reject data-entry outliers.
• Graceful fallback. If a state or locality has too few verified data points to be reliable, we fall back to the state-wide markup, then to a national figure (currently roughly 2.4× Medicare for BLS and 2.7× for ALS).
• It improves over time. Every new verified rate we add re-fits these multipliers, so estimates for unverified areas get sharper as our coverage grows.

The result is labeled clearly on every page: a single number with a source link means it's a verified local rate, while a range with a "Methodology" note means it's a modeled estimate. Treat estimates as a negotiating anchor, not an exact invoice — actual bills vary by provider.

Extending Our Reach: Crowdsourcing & Manual Research

While the CMS dataset provides the federal benchmark, we recognize that local rates—especially for "Treatment Without Transport" fees—are not always captured in federal schedules. To provide the most accurate picture, we supplement our database through two primary methods: