Various scoring scales have been introduced in the management of patients with multiple injuries and lower extremity injuries. Two scoring systems have been introduced to predict instability in distal radius fractures. The purpose of this investigation was to evaluate the accuracy of these two models in predicting instability. A prospective study of 105 consecutive patients sustaining unilateral closed distal radius fractures was performed. Two scoring systems--the MacKenney formula and the Adolphson formula--were used to calculate the probability of fracture instability on the basis of initial presentation and injury films. The predicted probability of instability calculated from both models was then compared with actual results of instability on the basis of specific radiographic criteria at follow-up. Final follow-up information was available on 80 patients. There were 44 unstable fractures and 36 stable fractures at final follow-up. Using the MacKenney formula, of the 38 fractures predicted to have a low probability of instability (Pinstability < 30%), 18 (47.4%) were found to be unstable. Using the Adolphson formula, of the 28 fractures predicted to have a low probability of instability (Pstability > 70%), 14 (50%) were actually unstable. Both scoring systems were found to underestimate the degree of fracture instability and to have a negative predictive value between 47 and 50% in a prospective series of patients. In fractures predicted to have a low probability of instability in both models, we found a poor correlation between predicted instability and actual instability. Our results demonstrate the limitations of two scoring systems in predicting fracture stability and in making clinical decisions on the basis of their results.