We have compared the Harris hip score with the Oxford hip score in a population of 358 patients (213 men and 145 women) aged between 19 to 74 years (median 55 years), after resurfacing arthroplasty of the hip (between September 1995 and October 2006) with a median follow-up of 6 years. The Oxford hip score was related to the age of the patient (Mann-Whitney test; p = 0.015), the hip lifetime (p = 0.030) and body mass index (p < 0.001). Correlation analysis indicated a good correlation between overall Harris and Oxford hip scores (Spearman's rank correlation = -0.70; p < 0.001). An analysis of correlations between individual items in the Oxford score and functional domains of the Harris score showed that the range of movement domain of the latter score was correlated with two items from the former score (-0.40 and -0.38; p < 0.001 and p < 0.001). Based on the correlation analysis, this study provides good evidence that the Oxford score can be substituted for the Harris score for long-term assessment of hip function, without significant loss of information.