Although extensive efforts have been made to optimize the safety of vectors for somatic gene transfer and somatic gene therapy, the safety of these agents must ultimately be assessed in clinical trials. A statistically significant assessment of safety will be complicated by the relatively small number of patients who will be enrolled in initial clinical trials, the need for long-term longitudinal follow-up of patients and perhaps their progeny, and the traditionally poor participation in long-term follow-up by many patients. This article reviews the potential risks of retroviral-mediated gene transfer, the statistical power required to assess the true incidence of potential complications, the number of patients who may participate in clinical trials involving retroviral vectors, and the factors that make thorough follow-up and uniform data ascertainment difficult. The role of the FDA in assessing the safety of retroviral vectors and the potential role of registries for patient tracking and data ascertainment are discussed.