Survivorship analysis, which is often encountered in the medical literature, is used to calculate the probability of a certain event, such as failure of a prosthesis, as a function of the time elapsed since an operation. Possible pitfalls in the use of this method are related to the size of the population of patients and the definition of how the outcome is measured. We studied the outcomes of 204 total knee arthroplasties in 165 patients, using six different end-points, in order to illustrate these problems. Survivorship estimates that are cited without confidence intervals have little clinical value.