Of the 141 osteoarthritic knees treated by tibial osteotomy at the Kotka Central Hospital from 1969 to 1974, 103 (ninety-two with a varus deformity and eleven with normal alignment) were followed for at least five years, the average follow-up being 6.9 years. In eighty-six knees (83.5 per cent) the result was good or fair. In the ninety-two knees with a varus deformity, the best results were achieved when the postoperative femorotibial angle was 167 to 175 degrees. In the eleven knees with normal preoperative alignment, the osteotomy was designed to shift the load to the compartment in which there was less osteoarthritis involvement. In nine of these eleven knees there was an average change in angulation (varus or valgus) of 5.3 degrees (range, 3 to 9 degrees) and the result was good or fair. All osteotomies united without reoperation. The incidence of delayed union was 3.6 per cent when the osteotomy was proximal and 14 per cent when it was distal to the tibial tubercle. The elimination of the tethering effect of the fibula by osteotomy or release of the proximal tibiofibular joint did not prevent the recurrence of varus deformity. At an average of 3.4 years after the osteotomy, deterioration was demonstrable in twenty-six knees in which the initial result had been good. This is a higher incidence of deterioration than in previously published series. In sixteen of the 103 knees, a total arthroplasty subsequently was performed because of a poor result, at an average of 7.6 years after the osteotomy.