Skip to main content
  • Dataset id: fda_drug
  • Jurisdiction: US · FDA
  • Date field: date_decision. The search’s top-level date_from and date_to filter on it.
  • Access: any API key.

​
Filters

Send these under filters.fda_drug in a search request. A filter not listed here, or a value of the wrong type, is a 400. The top-level date_from and date_to set date_decision_from and date_decision_to below, so give a date in one place only.
​
string | string[]
Exact match on the FDA application number; pass a string for one value or a list for OR-match. NDA/ANDA/BLA prefixes, spaces, and punctuation are stripped and digits zero-padded to 6 (‘ANDA 202121’ - ‘202121’, ‘4782’ - ‘004782’). The most precise filter — use it whenever a prior result, citation, or document URL already gives you the application number.
​
string
Filter by brand/product name. Example: ‘Paxlovid’.
​
string
Filter by active ingredient. Example: ‘nirmatrelvir’.
​
string
Keyword match on sponsor/company name (BM25 tokenized, case-insensitive). Example: ‘Pfizer’.
​
string
Filter by application type. NDA = New Drug Application (brand-name drugs). ANDA = Abbreviated NDA (generics). BLA = Biologics License Application.Values: NDA, ANDA, BLA.
​
string
Uppercased before matching. ORIG = original application. SUPPL = post-market supplement.Values: ORIG, SUPPL.
​
string | string[]
Chemistry classification code of the submission (uppercased). Pass a string for one value or a list for OR-match. Common values: ‘TYPE 1’ (NME), ‘TYPE 1/4’, ‘TYPE 2’, ‘TYPE 3’, ‘TYPE 4’, ‘TYPE 5’, ‘TYPE 6’, ‘TYPE 7’, ‘TYPE 8’, ‘TYPE 9’, ‘TYPE 10’, ‘BIOSIMILAR’, ‘BIOEQUIV’, ‘EFFICACY’, ‘LABELING’, ‘MANUF (CMC)’, ‘MEDGAS’, ‘REMS’. Combined variants also exist (e.g. ‘TYPE 2/3’, ‘TYPE 3/4’, ‘TYPE 4/5’) — pass them in a list when relevant. For BLA-vs-NDA filtering use the application_type field, not this one. For ‘novel drug’/NME questions, combine submission_type=‘ORIG’ with submission_class_code=[‘TYPE 1’,‘TYPE 1/4’].
​
string
FDA review track. PRIORITY = ~6-month review target. STANDARD = ~10-month target.Values: PRIORITY, STANDARD, UNKNOWN.
​
string
Current Drugs@FDA marketing status. Use for ‘is X still marketed?’ / ‘which approvals were discontinued?’. ‘marketed’ = currently on the market (Prescription OR Over-the-counter). ‘discontinued’ = withdrawn from marketing. ‘prescription’/‘over-the-counter’ = the specific Rx/OTC status. ‘tentative-approval’ = approved but never marketed. Matches if ANY of the application’s products has that status. Also available as a return column (lists all current statuses).Values: marketed, discontinued, prescription, over-the-counter, tentative-approval, for-further-manufacturing-use.
​
string | integer
label = prescribing information (concise, structured). review = FDA review documents (broadest content, dispersed across many pages). letter = approval/action letter (short). other = real documents FDA typed outside those three. An integer is treated as a raw ApplicationDocsTypeID (as returned in the document_type column). See the document_type CAVEAT in the tool description: FDA’s typing is unreliable — if an expected document doesn’t appear, re-run WITHOUT this filter.Values: label, review, letter, other.
​
string (date)
Inclusive lower bound on approval date (YYYY-MM-DD). Example: ‘2024-01-01’.
​
string (date)
Inclusive upper bound on approval date (YYYY-MM-DD). Example: ‘2024-12-31’.

​
Fields

Each result from this dataset has these in fields: active_ingredient, app_number, application_type, company, date_decision, document_type, drug_name, id, marketing_status, product_number, reference_drug, review_priority, submission_class_code, submission_class_code_description, submission_number, submission_type, url