Background: US FDA statistical and medical reviews of New Drug Applications (NDA reviews) are posted publicly on a website (www.drugs@fda.com). These reviews summarize results of clinical trials submitted by drug manufacturers in support of an application for approval of a new drug or a new indication for an already-approved drug. They often contain details of unpublished studies and additional unpublished results of published trials and may provide detailed assessments of the methodological quality of submitted trials. Objectives: To evaluate the completeness of the FDA website as a source of evidence for systematic reviews. Methods: We assessed the online availability of FDA reviews for 136 drugs included in 15 reports conducted for the Drug Effectiveness Review Project (DERP) from 2005 to March 2008. Additional analyses are underway. Results: Overall, NDA reviews were available for 52.2% of the drugs included in DERP reports (Table 1). The percentage of reviews available for different drug classes ranged from 14% to 100% (median 67%). More recently-approved drugs were more likely to have a review available; 86% of drugs approved between 1997 and 2006 had a review available versus 8% of drugs approved before 1997 (Figure 1). Since 43% of drugs included in the DERP reports were approved before 1997 (*Figure 2), the usefulness of the FDA website as a resource for comparative drug class reviews is limited. We also found that two-thirds of very newly-approved drugs (i.e., approved in 2007 or 2008) were missing review documents. There were approvals for 121 new or modified indications posted at FDA for 56 drugs in our sample (Table 1, column 5). Reviews of the data associated with these new indications were available for only 38%. Conclusions: The FDA website is a potentially valuable but incomplete source of data for systematic reviews of drugs approved in the US. For newer drugs, especially those approved after 1996, it is a useful resource. It would be more useful for comparative drug class reviews if more NDA reviews on older drugs were added. Systematic reviewers should consider adding this database to their search strategies.