We analysed two series of patients affected by unicompartmental arthrosis or axial malalignment of the knee treated with two different techniques of high tibial osteotomy. Forty-seven knees were treated with a closing wedge osteotomy (CWO) and 40 with an opening wedge osteotomy (OWO). The two groups were comparable with respect to age, gender and deformity. For each patient the patellar height was measured by Caton's method before surgery, and at the latest assessment (at least 1 year after operation). The correction rate for the two series was analysed to assess any possible correlation between the variation of the patellar height and the degree of correction of the knee axis. We concluded that a high tibial osteotomy modifies the patellar height and that this depends on the technique employed. Patellar 'lowering' occurred more often with OWO than with CWO and the latter also produced a high degree of patellar elevation.