We reviewed prospectively cemented stem revision in 106 patients with severe femoral endosteal bone lysis without infection. Bone grafts were not used in any of the patients. The minimum follow-up was three years (mean 6 years 4 months). At the last review 76.4% of the patients were free of pain and 17.9% had only mild or occasional discomfort; radiographs showed well-fixed stable stems in 101 (95.3%). An intramedullary cement plug was used at revision to improve stability in 97.7%. There was new endosteal osteolysis after revision in 17 patients; only two had severe changes. Seven hips (6.6%) required a second revision; only four of these (3.8%) were for stem loosening, Survivorship of the revised stem, using radiological evidence of stem loosening as the end point, was 95.8% at seven years. The results of stem revision arthroplasty using cement in the presence of massive endosteal cavitation are satisfactory.