The validity of observational studies is sometimes questioned because of the limitations of non-randomly assigned controls, various biases such as channeling bias, confounding by indication, and other pitfalls. Yet, (post-marketing) observational data can provide important information regarding not only drug safety but also the effectiveness and appropriate use of agents in the real world, outside of clinical trials. Observational studies also provide data regarding the wider value of these agents in terms of, for example, reducing the need for surgical procedures, reducing absenteeism and increasing productivity. Importantly, data from some observational registry studies have subsequently been confirmed by clinical trials, supporting the overall validity of the registry-based approach. Observational studies also allow measures such as health assessment questionnaire scores, disease activity scores, and glucocorticoid use over time to be monitored for longer periods. Furthermore, observational data in real, less strictly selected patients without the constraints of formal study populations may produce findings not observed in clinical trials but that warrant further investigation in a controlled trial environment. For example, recent data from the Stockholm tumor necrosis factor follow-up registry in Sweden showed increases in the time people worked after initiation of biologics that, surprisingly, continued into the fourth and fifth years of treatment--a finding not observed with standardized outcomes. Observational studies are truly an underappreciated and valuable source of data on the real value of anti-rheumatic therapies, and these data are essential for making sound decisions regarding coverage and reimbursement.