Assessing safety is important to evaluating new medications. In accordance with the Food and Drug Administration's guidance published in 2008, many studies with cardiovascular outcomes have been conducted to evaluate the safety of newly approved medication. This paper discusses two approaches to analysis of safety data: intention-to-treat (ITT) and on-treatment analysis. We refer to principles of the design of randomized clinical trial, from the perspectives of patients and from data analysis. On-treatment analysis violates randomization and is often not well defined. Moreover, because the typical on-treatment analysis ignores the reason participants in clinical trials stop treatment, on-treatment analyses often leads to biased estimates of risk. Pure intention-to-treat analyses that capture all events from randomization to the end of the trial have limitations, but we recommend that for major safety outcomes where cause cannot be attributed to the individual case, ITT analysis should be performed in the framework of benefit-risk evaluation. Analyses of safety, as well as efficacy, should be tailored to the specific question being asked and with specific study design in mind.