We have conducted a systematic review of the scientific literature concerning outcome and clinical effectiveness of prostheses used for primary total hip replacement (THR) in Norway. The study is based on two Health Technology Assessment reports from the UK (Faulkner et al. 1998, Fitzpatrick et al. 1998), reviewing the literature from 1980 to 1995. Using a similar search strategy, we have evaluated the literature from 1996 through 2000. We included 129 scientific and medical publications which were assessed according to a specific appraisal protocol. The majority (72%) were observational studies, whereas only 9% were randomized studies. We could not retrieve any peer-reviewed documentation for one third of the implants. The Charnley prosthesis had by far the best and most comprehensive evidence base with better than 90% implant survival after about 10 years. Survival of the Charnley prosthesis declines by about 10% during each of the two following decades. Except for the Charnley and Lubinus IP, no other prosthesis on the market in Norway has given long-term results (> 15 years). 5 other cemented implants have given comparable results at about 10 years of follow-up. Some uncemented stems have shown promising medium-term outcome, but no combination of uncemented cup and stem fulfilled the benchmark criterion of > or = 90% implant survival at 10 years, which we propose as a minimum requirement for unrestricted clinical use for prostheses used in primary THR. New or undocumented implants should be introduced through a four-step model including preclinical testing, small series evaluated by radiosterometry, randomized clinical trial involving comparison with a well-documented prosthesis, and finally, surveillance of clinical use through registers.