BACKGROUND: The purpose of the present study was to compare the results associated with two different surface finishes (bead-blasted and grit-blasted) for Iowa femoral components used in total hip arthroplasties performed between January 1979 and June 1991 in patients who were less than fifty years old. METHODS: Between January 1979 and December 1985, thirty-six primary total hip replacements were performed in twenty-five patients with insertion of a bead-blasted Iowa femoral component (average surface roughness, 0.8 micrometer) with cement and insertion of a titanium-backed acetabular component (thirty-five hips) or non-metal-backed acetabular component (one hip) with cement. Between January 1986 and June 1991, forty-five primary total hip replacements were performed in thirty-seven patients with use of a precoated grit-blasted Iowa femoral component (average surface roughness, 2.1 micrometers) and a Harris-Galante-I porous ingrowth acetabular component (forty-one hips) or an Osteonics component (four hips). The only change in the design of the femoral component (other than the surface finish) between the two consecutive series was the addition of polymethylmethacrylate precoating to the proximal third of all forty-five grit-blasted stems and modularity of the femoral head of the last eight grit-blasted stems. RESULTS: No hip was lost to follow-up. The duration of radiographic follow-up of the hips treated with the bead-blasted component averaged 11.3 years (range, ten to sixteen years) postoperatively, whereas that of the hips treated with the precoated grit-blasted component averaged 8.2 years (range, five to eleven years). Two (6 percent) of the bead-blasted femoral components were revised because of aseptic loosening compared with eight (18 percent) of the precoated grit-blasted components. Four bead-blasted components (11 percent) were either radiographically loose or were revised because of loosening compared with eleven precoated grit-blasted components (24 percent). Kaplan-Meier survivorship curves evaluated with log-rank analysis revealed that the bead-blasted Iowa femoral components were revised because of aseptic loosening (p = 0.0184) and were radiographically loose (p = 0.0068) less often than the precoated grit-blasted Iowa femoral components. CONCLUSIONS: The findings of the present study have led the senior ones of us to resume using femoral components with a polished surface and fixed with cement.