Introduction: In this paper, the Gait Deviation Index (GDI) was used as a convenient method to evaluate pre-to-postoperative gait pattern changes after total hip arthroplasty and identify factors which might be predictive of outcome. Design: Three-dimensional gait data from a randomized clinical trial was used to determine changes in gait quality in participants walking at self-selected speed. Upon completion of the first assessment, the participants were randomly assigned to either resurfacing hip arthroplasty or conventional hip arthroplasty. The outcome was changes in overall gait ‘quality’ measured with GDI during the 6-month post-surgery follow-up period. Results: 38 participants with severe unilateral primary hip osteoarthritis took part in the trial. We found no difference in change scores between the two treatment groups; 1.9 [95%CI: 0.3 to 4.0] or between change scores for the non-operated and the operated limbs; 0.3 [95%CI: 2.3 to 1.7]. However, the score for the two groups (pooled data) improved after surgery by 4.4 [95%CI: 1.8–7.0]. The single level regression analysis identified the preoperative GDI score as a strong predictor of outcome (p < 0.001). Conclusion: Six months after surgery, there was no additional effect of resurfacing hip arthroplasty on GDI scores compared with conventional hip arthroplasty. Participants with the most pathological preoperative gait pattern improved the most. The GDI increased, which indicates an overall improvement in gait pathology after surgery. Trial registration: NCT01229293