OBJECTIVE The aim of this study was to analyze the setting for dislocations and redislocations after primary and revision total hip arthroplasty (THA), identify risk factors and optimize treatment. METHODS This study included 56 patients with a dislocated hip following THA (n=5,205) between 1984 and 2005 and a matched control group (n=55). Hospital charts and radiographs of all patients in both groups were analyzed. Thirty-one patients in the study group were followed both clinically and radiologically. RESULTS The dislocation rate after primary THA was 1.1% (56/5,205) and the redislocation rate after a first occurrence was as high as 39%. There was a positive correlation between the time intervals from the surgery to first dislocation and from the first dislocation to second dislocation (r=0.4). Most of the primary dislocations occurred within a short period of time after surgery, thus favoring consecutive dislocations. Female gender, as well as revision arthroplasty, was associated with a higher incidence of dislocations. No relation was found between the orientation of the acetabular cup and dislocation. CONCLUSION To prevent redislocations after the first occurrence, we suggest thorough evaluation of possible technical faults which should be addressed surgically. Considering the high redislocation rate, we also advocate a stringent conservative treatment regime especially after the first THA dislocation.