OBJECTIVE: To determine how long it takes after a new drug is marketed to establish whether or not its use by pregnant women is likely to pose a substantial teratogenic risk. METHODS: We used standard clinical teratology resources to assess the teratogenic risks in human pregnancy of therapeutic treatment with 468 drugs approved by the US Food and Drug Administration between 1980 and 2000. The teratogenic risk of each treatment was classified using the current online version of TERIS into one of three categories: 1) no risk, minimal risk, or unlikely to produce an increased risk; 2) associated with a small, moderate, or high risk; or 3) risk undetermined. RESULTS: We found that the teratogenic risk in human pregnancy was still undetermined for 91.2% of drug treatments approved in the United States between 1980 and 2000. The proportion of treatments classified as having an “undetermined” teratogenic risk was more than 80% for drugs approved for marketing 0–4, 5–9, 10–14, or 15–20 years ago, but the highest proportion of drugs with an “undetermined” teratogenic risk was found among those approved 15–20 years ago. The agreement between TERIS risk ratings and Food and Drug Administration Use-in-Pregnancy Categories for 163 drugs that had been assessed by both systems was poor (κ ± standard error = 0.082 ± 0.042). CONCLUSION: We conclude that inadequate information is available for pregnant women and their physicians to determine whether the benefits exceed the teratogenic risks for most drug treatments introduced in the past 20 years.