Labral tears are commonly associated with femoroacetabular impingement. We reviewed 151 patients (156 hips) with femoroacetabular impingement and labral tears who had been treated arthroscopically. These were subdivided into those who had undergone a labral repair (group 1) and those who had undergone resection of the labrum (group 2). In order to ensure the groups were suitably matched for comparison of treatment effects, patients with advanced degenerative changes (Tönnis grade > 2, lateral sourcil height < 2 mm and Outerbridge grade 4 changes in the weight-bearing area of the femoral head) were excluded, leaving 96 patients (101 hips) in the study. At a mean follow-up of 2.44 years (2 to 4), the mean modified Harris hip score in the labral repair group (group 1, 69 hips) improved from 60.2 (24 to 85) pre-operatively to 93.6 (55 to 100), and in the labral resection group (group 2, 32 hips) from 62.8 (29 to 96) pre-operatively to 88.8 (35 to 100). The mean modified Harris hip score in the labral repair group was 7.3 points greater than in the resection group (p = 0.036, 95% confidence interval 0.51 to 14.09). Labral detachments were found more frequently in the labral repair group and labral flap tears in the resection group. No patient in our study group required a subsequent hip replacement during the period of follow-up. This study shows that patients without advanced degenerative changes in the hip can achieve significant improvement in their symptoms after arthroscopic treatment of femoroacetabular impingement. Where appropriate, labral repair provides a superior result to labral resection.