From 1983 to 1985 we performed 114 primary hip replacements in 108 consecutive osteoarthritic patients using a non-cemented RM isoelastic femoral stem. After a mean follow-up of 8.2 years, ten patients had died, 11 hips had been revised, six patients had been lost to follow-up and two had been excluded due to severe general illnesses. Of 85 arthroplasties (in 79 patients) 14 could not be assessed because of other illness or disability. The 71 remaining were reviewed by questionnaire and radiography; an excellent or good overall functional result was found in 31, 16 were fair and 24 were poor. Radiographically, 21 of 71 stems were judged to be loose and ten showed osteolytic foci, six of these without obvious loosening. We conclude that the isoelastic RM stem shows a high rate of loosening, but that this is not always associated with a poor subjective result. Regular radiographic review is necessary. The results are worse than those reported for other uncemented stems and for cemented stems.