The present investigation was undertaken to explore changes over time in the design and reporting of trials of second-line drugs in rheumatoid arthritis, in the characteristics of patients included in the trials, and in the sources of funding. We studied 105 trials of second-line agents for the treatment of rheumatoid arthritis, including placebo-controlled and comparative trials involving 8 different agents. Three time periods, 1945-1969, 1970-1979, and 1980-1989, were compared. We found little change in the standards for reporting on the design of trials or for reporting information on patient dropout and drug side effects, some increase in the complexity of the statistical methods used, but no increase in the use of power analysis (reported in only 13% of trials). The average age of patients in clinical trials has increased. In recent years, there has been a substantial shift from placebo-controlled to comparative trials, and it is increasingly common for trials to be financially supported by pharmaceutical companies. The possible effects of secular trends should be considered when combining or comparing results of trials conducted in different years.