Generic prices and the number of competitors, from FDA's curve to GxRadar's basket

When a brand-name drug faced its first generic between 2015 and 2017, the lone generic maker's price came in 39 percent below the brand's, measured at the manufacturer level. With four makers the gap was 79 percent. With six or more, generic prices sat more than 95 percent below the old brand price. Those are the medians in FDA's December 2019 analysis¹ of 181 products, one of the studies the agency lists on its generic competition and drug prices page².

GxRadar's own basket of 47 common generics sits at the far end of that curve. Every one of them has at least six approved generic applications with a product still marketed, by our count of FDA's Drugs@FDA records. Famotidine has 60. Warfarin has six.

This piece sets the federal research on competitor counts next to those application counts and to what pharmacies paid for each drug, as measured by the federal NADAC survey. In crowded, long-established generic markets the link between more approvals and lower prices turns out to be weak. The table and the caveats below explain why.

What FDA's 2019 study measured

FDA economists studied products whose first generic entered between January 2015 and December 2017 and compared each month's generic price with the brand's price before entry. They used two price measures. One was the average manufacturer price that drugmakers report to Medicaid. The other was the invoice price pharmacies pay wholesalers. The final sample covered 181 products and 3,688 product-months, and the count of producers included authorized generics, which are the brand's own drug sold under a generic label (GxRadar explains authorized generics separately).

Medians fall with nearly each added producer. The one reversal, from seven makers to eight, rests on just 49 product-months. Across all competition levels, the median generic was 40 percent of the brand price on the manufacturer measure and 49 percent on invoices. FDA attributed most of the difference between the two measures to wholesaler markups. It also noted that 61 percent of the product-months in its sample had only one or two generic competitors, so most observations sit at the expensive end of the table. GxRadar's earlier piece on why generic prices fall as competitors enter walks through the mechanics.

Medicare data point the same way

HHS's Office of the Assistant Secretary for Planning and Evaluation (ASPE) reached a similar result with Medicare Part D claims. Its January 2025 issue brief³ looked at drugs whose first generic arrived in 2007 to 2011, 2012 to 2015 and 2016 to 2019, and measured prices three years after entry. Prices were about 20 percent lower in markets with about three competitors and 70 to 80 percent lower in markets with 10 or more, relative to the price before generics.

ASPE also found that most new generic markets stay small. Sixty percent of the drugs in its study had fewer than 5,000 Medicare users a month, and those markets averaged about three labelers. Markets with 15,000 or more monthly users averaged close to 10. The size of the market, in other words, helps decide how many companies show up.

FDA's most recent savings estimate points the same direction. The agency fully approved 773 generic applications in 2023 and estimated $18.6 billion in savings over the following 12 months, according to its November 2025 report⁴. First generics accounted for $2.4 billion of that, or 13 percent. "Nine of the top ten drugs by savings in 2023 were not first generics," the report said, because later entrants into thin markets also pushed prices down.

FDA's review rules reflect this. Under its prioritization policy, MAPP 5240.3⁵, an application can qualify for priority review when there are "not more than three approved drug products" listed in the Orange Book⁷ and no blocking patents or exclusivities. A separate competitive generic therapy⁶ pathway, created by Congress in 2017, applies when there is not more than one approved drug in the Orange Book's active section. GxRadar covers that program and the drugs that end up with only one maker.

What GxRadar's application counts show

For each drug in our price basket, GxRadar counts the abbreviated new drug applications (ANDAs) in FDA's Drugs@FDA database, read through the openFDA drugsfda endpoint⁸, whose active ingredient matches the drug. A "marketed" application has at least one product FDA lists as prescription or over-the-counter rather than discontinued. The counts cover every strength and dosage form, so a drug's count can include injectable, liquid or over-the-counter versions as well as the tablet or capsule whose price we track. Across the 47 drugs, the database holds 1,762 such applications, and 1,040 still list a marketed product.

Prices come from CMS's National Average Drug Acquisition Cost files⁹, monthly averages through October 2026. NADAC is a survey of the invoice prices retail pharmacies pay, and under the CMS methodology¹¹, discounts or rebates not shown on the invoice are not counted. It is not what a customer pays at the counter. The table groups the basket by the number of marketed applications.

Drugs with the most marketed applications fell furthest over five years, a median of 35 percent, against 17 percent for the least crowded group. The pattern is real but loose. Ranked drug by drug, the number of marketed applications explains little of the variation in five-year price change, and each group contains drugs that rose. Of the seven basket drugs that cost pharmacies more in October 2026 than in January 2014, two sit in the most crowded group.

Fill sizes in the price column come from GxRadar's analysis of CMS's State Drug Utilization Data for 2025¹⁰, which shows how Medicaid prescriptions for each drug were typically dispensed. They describe dispensing patterns, not dosing. GxRadar explains how state Medicaid programs pay for these fills separately.

Why application counts are a rough proxy

An approved ANDA is permission to sell, not proof of sales. One company can hold several applications for the same ingredient, applications change hands, and a "marketed" flag in Drugs@FDA can lag behind a product that has quietly left the market. FDA's 2019 study¹ counted producers actually selling in each month, and ASPE³ counted labelers on Medicare claims. Neither is the same as an application tally.

The basket also sits on the flat part of the curve. FDA's medians show that by six or more producers, generic prices are already a few percent of the brand's. Going from 20 approved applications to 60 adds little room for further declines when 30 tablets of famotidine cost pharmacies 86 cents.

Furosemide shows how a crowded market can still move up. It has 22 applications with a marketed product, and its NADAC for 30 tablets rose from 31 cents in January 2014 to 87 cents in October 2026. That is a 180 percent increase on a price that remains under a dollar. GxRadar tracks it and the other six risers in older generics that cost more than in 2014. Warfarin, with the fewest marketed applications in the basket, cost $2.67 for 30 tablets in October, 9 percent above January 2014.

Readers can check any basket drug's application count on its GxRadar drug page and FDA's yearly totals on the approvals page. The generic drug user fee program, which funds much of FDA's review work, is authorized through September 2027, the ASPE brief³ notes. It is up for renewal next year, and the review timelines it sets govern how quickly new competitors can reach markets like these.