We studied 122 controlled clinical trials and observational studies of second line therapy that involved 16,071 patients. The mean disease duration was 7.61 years at study entry. Controlled clinical trials were inherently short term, and 90.5% of patients were followed for 1 year or less (mean 8.8 months). The mean followup of observational studies was 31 months. Outcome assessments that included functional measurements were rare in either study type, as were considerations of socioeconomic factors. Except for methotrexate, which was used longer, half of the studies indicated discontinuation of therapy after 1.5 years. Good retention rates in studies of 3 to 12 months were not representative of longterm results, but controlled clinical trials and observational studies were similar as to retention during the first treatment year. Observational studies following controlled clinical trials can yield important information about RA treatment effectiveness not available from controlled clinical trials alone.