Proximal tibial osteotomy is commonly performed for osteoarthritis of the knee with deformity. The results of 105 dome osteotomies have been reviewed at a minimum follow-up of one year and an average of 4.8 years. Before operation all the knees were painful, 50.5% severely; a further 45.7% disturbed sleep at night. At review 15.2% of knees were free of pain and 60% had only slight pain which did not restrict activity. The preoperative range of movement was maintained and there was only a slight tendency for radiological changes to progress, with actual improvement in some cases. No correlation was found between the correction of deformity to physiological valgus and the result. We cannot explain why tibial osteotomy produces such useful and sustained pain relief.