Built on WIX Headless
Pharmacy Market Map

Methodology and limits

Every number on this site comes from public data. This page says where each one comes from, how the signal is calculated, and where the data is weak.

What the signal means

For each county, the signal compares three things with every other county: residents per community pharmacy, the share of residents aged 65 and over, and median household income. It is a comparison between counties. It does not assess a specific location, rents, insurer contracts or mail-order competition, and it is not investment advice.

Sources

DataSourceDate
Pharmacies in Medicare Part D retail networksCMS, Monthly Prescription Drug Plan Formulary and Pharmacy Network InformationOctober 2026
Name, address and type of each pharmacyCMS, NPI registry (NPPES) full fileNovember 2026
Residents, residents 65 and over, median household income, income per personU.S. Census Bureau, American Community Survey 5-year estimates2019 to 2023
Pharmacy and drug store establishments (shown for reference)U.S. Census Bureau, County Business Patterns, NAICS 4461102023
County for each addressU.S. Census Bureau geocoder; ZIP code to county where an address could not be geocoded (about 11% of addresses)2026
County and state shapes on the mapsU.S. Census Bureau, cartographic boundary files2023

The maps are colored by residents per community pharmacy, in five bands that each hold a fifth of U.S. counties. A dark county has many residents for each pharmacy. That is a ratio, and it can be dark without earning the underserved signal, which also has to survive the error test described below.

How pharmacies are counted

We start from pharmacies marked as retail in at least one Medicare Part D plan network, match them to the NPI registry, drop deactivated records, and remove duplicates by address. A pharmacy counts as a community pharmacy if its NPI taxonomy is retail or general pharmacy in the primary position, or retail in any position. Military pharmacies, dispensaries that are not pharmacies, records registered as individual pharmacists and mail-order pharmacies are excluded. Long-term care, clinic, infusion, specialty, compounding, institutional and Indian Health Service pharmacies are counted separately and do not enter the signal.

Residents per pharmacy is our own calculation: residents divided by community pharmacies. Medians for states and for the country are medians across counties.

The signal formula

Each county gets a score from 0 to 100: its percentile among counties for residents per community pharmacy (60% of the score), share of residents 65 and over (25%), and median household income (15%). A score of 70 or more points to an underserved market. A score below 40 points to a crowded one. The weights and cut-offs are our judgment and have no external validation.

A county is shown as underserved only if it would still be underserved with 15% more pharmacies than we counted (or one more), and as crowded only if it would still be crowded with 15% fewer (or one fewer). Everything else is shown as no clear signal. Counties with fewer than 10,000 residents get no signal. Counties where our data shows no community pharmacy get no signal either, because we cannot tell a real gap from a gap in the data.

How wrong the count can be

We compared our community pharmacy count with official state license lists. It was 3% below the list in Connecticut, 10% below in the District of Columbia and 13% below in Delaware. The count tends to run low. That measured error is the reason for the 15% test above.

Known limits

Data build: 2026-11-21.