Background: Cementless femoral stems in total hip arthroplasty provide dependable clinical and radiographic results in the treatment of osteoarthritis. Stem length might affect the preservation of proximal bone stock and stability. We hypothesized that a shorter stem decreases proximal bone loss without affecting implant stability. Methods: We randomly assigned 60 patients aged between 50 and 70 years to either a standard cementless femoral stem or a 35-mm shorter version. Patients were followed with dual-energy X-ray absorptiometry, radiostereometric analysis, Harris hip score, Western Ontario and McMaster Universities Osteoarthritis Index score, and clinical follow-up at 3, 6, 12, and 24 months. The trial is registered on ClinicalTrials. gov/. Results: After 24 months, short stems had on average 3.8% (95% confidence interval, 1.2%-8.9%) more bone loss in zone 1 compared to standard stems (P - .14). In zone 7, the bone loss was on average 6.5% (95% confidence interval, 6.6%-19.7%) higher compared to standard stems (P - .33). After 24 months, standard stems had migrated 0.93 mm (range, 0.25-4.66 mm) and short stems 0.93 mm (range, 0.17-2.96 mm; Student t-test after log transformation, P - .3). Patient-reported outcome measures were similar in both groups. One patient in the standard stem group was diagnosed with infection, one with a posterior dislocation, and one with a deep venous thrombosis. No stems were revised. Conclusion: There were no statistically significant differences in periprosthetic bone loss or fixation between the stems at 24 months. (C) 2016 Elsevier Inc. All rights reserved.