Extended-release, delayed-release and immediate-release pills, and how FDA judges their generics

Since June 2012, FDA's enforcement records have logged 474 recall records for extended-release tablets and capsules. In 182 of them, about 38 percent, the stated reason was a failed dissolution test, the lab check of how fast a pill releases its drug. For tablets and capsules not identified as extended- or delayed-release, the rate was 4.5 percent. Those figures come from GxRadar's analysis of FDA's openFDA enforcement data¹ through Oct. 28, 2026, sorting each record by the words in its product description.

That gap says something about what an extended-release pill is. Its whole job is to release a drug slowly and on schedule, so a lot that releases too fast or too slow fails the test that matters most. The same timing is what FDA checks hardest when it approves a generic copy, and it decides how the Orange Book rates one product as a substitute for another.

Three ways a pill releases its drug

FDA sorts oral products into immediate-release and modified-release, and splits modified-release into two kinds. Its 2022 bioavailability guidance² and its 2026 generic bioequivalence guidance³ give the working definitions.

An immediate-release product is the conventional pill, which lets its drug dissolve and absorb without any built-in delay. An extended-release product "allows a reduction in the dosing frequency and reduces fluctuations in plasma concentrations" compared with an immediate-release form, in the words of the 2026 guidance. It can be a capsule, tablet, granules, pellets or a suspension. FDA uses "extended," "controlled" and "sustained" interchangeably.

A delayed-release product releases its drug "at a time later than immediately after administration." Typically an enteric coating holds the drug until the pill has passed through the acid of the stomach. The 2022 guidance notes that many delayed-release products "behave like IR products after accounting for the delay."

Several drugs on GxRadar's list show the split. Metoprolol succinate is sold as an extended-release tablet, while metoprolol tartrate is immediate-release. Omeprazole, pantoprazole and duloxetine are delayed-release.

The letters on the label

Official dosage-form names spell the release type out in full. The Orange Book lists products as "TABLET, EXTENDED RELEASE," "CAPSULE, DELAYED RELEASE" and so on. Generic product names usually spell out the form too, as in "Venlafaxine Hydrochloride, extended-release capsules, USP," the wording on a 2024 recall record. Brand names often add a short modifier instead, and those letters follow no single rule.

A recommendation from the National Coordinating Council for Medication Error Reporting and Prevention⁶ says modifiers "such as LA, CR, CD, XL and SR" have "no standard interpretation that would give a clear meaning to indicate release properties or dosing frequency." The council notes that USP's general chapter on dosage forms recognizes two categories of modified release, delayed and extended, and lists errors such as confusing "Drug Name 120 mg" with "Drug Name ER 120 mg," a cousin of the look-alike, sound-alike name problem. The full dosage-form name is the more reliable signal; GxRadar's guide to reading a prescription label shows where it appears.

How FDA judges a generic of each

Every generic must be bioequivalent to its reference product, meaning the rate and extent of absorption fall within the limits described in GxRadar's explainer on the 80-125 rule. What changes by release type is how many studies FDA expects and under what conditions. FDA finalized its ANDA bioequivalence guidance⁴ in May 2026, replacing a 2021 draft, according to its Federal Register notice⁵.

The fed study is the main difference. For all orally administered modified-release products, the guidance says³, FDA recommends studies "under both fasting and fed conditions, irrespective of the RLD labeling regarding food intake." For most immediate-release products, the guidance says, "single-dose BE studies conducted under fasting conditions typically provide greater discrimination" between test and reference, so one fasting study is enough. Multiple-dose studies are "generally not recommended" for modified-release generics, because single-dose studies are more sensitive to differences in release.

Alcohol is a specific worry. The 2026 guidance warns that a modified-release formulation can "lose its modified-release characteristics" with alcohol, "leading to a more rapid drug release." Its advice is that sponsors of certain products test release in media with various alcohol concentrations, and FDA's product-specific guidances say which products need it.

Brand-name extended-release drugs face their own rules. Under 21 CFR 320.25(f)⁷, as the 2022 guidance² summarizes it, an extended-release study must show that the product meets its release claims, that its profile "rules out the occurrence of any dose dumping," and that it performs consistently from one dosage unit to the next.

What the Orange Book codes say

The Orange Book reserves a code for extended-release products. FDA's 2026 Orange Book⁸ keeps the BC code for extended-release forms and explains that the agency does not consider different extended-release products with the same ingredient and strength therapeutically equivalent "unless equivalence between individual products in both rate and extent has been specifically demonstrated." Products with that evidence are coded AB, the rating covered in GxRadar's guide to therapeutic equivalence codes.

In practice nearly all extended-release generics now carry the AB rating. GxRadar counted marketed generic listings for oral tablets and capsules in FDA's October 2026 data files⁹.

Of the 116 extended-release listings with no code at all, 69 were over-the-counter products. Only one listing in the entire October file carried a BC code, and four extended-release generic listings carried BX, the code FDA uses when the data are "insufficient to determine therapeutic equivalence."

One wrinkle matters for extended-release drugs. When two brand extended-release products of the same ingredient are not bioequivalent to each other, FDA gives them different three-character codes. The Orange Book's own example is nifedipine extended-release tablets, where two brand products carry AB1 and AB2 and each generic is matched to one of them. A generic coded AB2 is considered equivalent only to the AB2 product. A pharmacist can say which reference product a dispensed generic is rated against.

Prices by release type

Federal survey data show what pharmacies have paid for products of each type, and how far those prices have fallen. These are CMS NADAC¹⁰ figures, the average price pharmacies paid, for the generic products on our price pages, from November 2013 to the October 2026 files¹¹. They compare different strengths and say nothing about which form suits a given patient.

Whether an extended-release or immediate-release version fits a particular prescription is a question for the prescriber and pharmacist. The Orange Book treats products as therapeutic equivalents only when they are pharmaceutical equivalents "in identical dosage forms," so an immediate-release generic is never rated as a substitute for an extended-release product. Recall notices for either type are listed on GxRadar's recalls page. The next changes to watch are in FDA's product-specific guidances, which the May 2026 guidance points applicants to for partial AUC, alcohol testing and dissolution media drug by drug.