From 1975 to 1986, 102 high tibial osteotomies for varus gonarthrosis were performed in 99 patients. Fifty-eight patients (60 knees) were reviewed at an average followup of 15 years (range, 10-21 years); of the remaining 41 patients, seven had a knee replacement, 18 had died, and 16 were lost to followup. The results, assessed according to the scoring system of the Hospital for Special Surgery and including the seven patients who underwent a knee replacement, were excellent or good in 37 (55%) knees and fair or poor in 30 (45%). Twenty-six patients of the current study previously were reviewed in 1986, with an average followup of 8 years, using the same clinical and radiographic criteria. In this group of 26 patients, excellent and good results decreased from 73% in 1986 to 46% in 1996. The knees in these 26 patients with a followup greater than 15 years had a statistically significant higher percentage of fair and poor results. No statistically significant differences in the results were found according to the amount of correction. Radiographic controls at followup were available for 45 of the 60 knees; a loss of correction greater than 5 degrees was observed in 11 knees. The results of this long-term followup study show that high tibial osteotomy for gonarthrosis allows a long period (range, 10-15 years) of relief of pain, good range of motion, and function in a large number of patients. Results tend to deteriorate with time, particularly after 15 years.