Reporting of adverse events (AEs) in randomized clinical trials (RCTs) is often lacking and with limited application in the real world, as RCTs are of short duration, include small numbers of patients, and are selective for subjects lacking in comorbid conditions. It is not surprising that new and unexpected safety concerns emerge with any new drug after it has been launched and used by many more patients. Part of the problem is inherent to the way safety data are reported in RCTs. This article focuses on some of the shortcomings of AE reporting in RCTs, especially those involving tumor necrosis factor (TNF) inhibitors. Discussion focuses on reporting of "time-to-event" issues, use of standardized incidence ratios for comparison to normal population or disease controls, use of "patient-years" when reporting AEs, and the problem of adequate sample size and power calculations that are lacking in safety outcome data trials.