Unicompartmental knee arthroplasty has been controversial since its introduction in the early 1970s. Refinements were made in patient selection, surgical technique, and prosthetic design. Ten-year follow up studies were reported that showed survivorship was slightly less than that reported for total knee arthroplasty but acceptable considering the theoretically conservative nature of unicompartmental surgery. Unicondylar knee arthroplasty can be an attractive alternative to osteotomy or total knee arthroplasty especially some middle-aged women. Approximately all studies with followups of 10 years or greater show that unicompartmental knee arthroplasty will have inferior survivorship to total knee arthroplasty whether from loosening, prosthetic wear, or secondary degeneration of the opposite compartment in the second decade. Recently there has been a resurgence of interest in doing unicompartmental knee arthroplasty, which was encouraged by easier recuperation, decreased hospital stays, and good functional results. Before doing a unicompartmental arthroplasty, the surgeon should answer four important questions: Is the disease truly unicompartmental? Can this be determined on a clinical examination and standard radiographs, or are more sophisticated studies such as a bone scan or an arthroscopy required? Second, if the patient does have unicompartmental disease are there any specific contraindications to the surgery? What are the limits of fixed deformity in varus or flexion that can be corrected by a unicompartmental replacement? Overcorrection of angular deformities has in the past led to increased wear of the opposite compartment. Therefore, how much should the knee be corrected? What is the minimal polyethylene thickness that is permissible?