The purpose of the present study was to review the early results of periacetabular osteotomy in the initial group of patients undergoing this procedure at the authors' institution. The first 21 hips in 19 patients with greater than 2 years followup, which represents the learning curve with this operation, were reviewed retrospectively. There were 14 females and five males with an average age of 21 years (range, 17-43 years). Intertrochanteric osteotomy was performed simultaneously on four patients with coxa valga and inadequate correction with periacetabular osteotomy alone. At an average of 38 months of followup (range, 24-52 months), the Mayo hip scores improved from an average of 46 points (range, 34-58 points) to an average of 68 points (range, 42-80 points). Hip range of motion declined slightly in all three arcs of motion. The lateral center edge angle of Wiberg improved from an average of 2 degrees to an average of 24 degrees. The loading zone angle (Tönnis) improved from an average of 24 degrees to an average of 11 degrees. The anterior center edge angle of Lequesne improved from an average of -6 degrees to an average of 38 degrees. Complications included two peroneal palsies, both of which resolved completely; three ischial fractures that healed uneventfully; three asymptomatic pubic nonunions; and asymptomatic heterotopic ossification in five patients. One patient underwent subsequent total hip arthroplasty for progressive arthritis and pain. Another patient required intertrochanteric osteotomy at a later date. The early results in this initial group of patients treated with periacetabular osteotomy show reliable radiographic correction of deformity and improved function with an acceptable complication rate. Patients should be counseled carefully about possible loss of motion postoperatively. Additional study is necessary to assess the long term results of this procedure.