Background and purpose — It is believed that in uncemented primary total hip arthroplasty (THA) the anchorage of the stem is dependent on the level of bone mineral density (BMD) of the femoral bone. This is one of the reasons for the widely accepted agreement that a cemented solution should be selected for people with osteoporosis or age > 75 years. We evaluated whether preoperative BMD of the femur bone is related to femoral stem migration in uncemented THA. Patients and methods — We enrolled 62 patients (mean age 64 years (range 49–74), 34 males) scheduled for an uncemented THA. Before surgery we undertook DEXA scans of the proximal femur including calculation of the T- and Z-scores for the femoral neck. Evaluation of stem migration by radiostereometric analysis (RSA) was performed with 24 months of follow-up. In 56 patients both preoperative DEXA data and RSA data were available with 24 months of follow-up. Results — None of the patients had a T-score below –2.5. We found no statistically significant relationship between preoperative BMD and femoral stem subsidence after 3 or 24 months. When comparing the average femoral stem subsidence between 2 groups with T-score > –1 and T-score ≤ –1, respectively, we found no statistically significant difference after either 3 or 24 months when measured with RSA. Interpretation — In a cohort of people ≤ 75 years of age and with local femur T-score > –2.5 we found no relationship between preoperative BMD and postoperative femoral stem subsidence of a cementless THA.