We studied whether acetabular bone mineral density (BMD) is better preserved after resurfacing hip arthroplasty (RHA) versus small diameter metal-on-metal total hip arthroplasty (THA). This randomised controlled trial included 82 patients. BMD was measured in 5 periprosthetic regions of interest (ROI) with dual-energy absorptiometry (DEXA) preoperatively, at 3 and 6 months, 1, 2, 3 and 5 years postoperative. 34 RHA and 26 THA had a complete 5 years follow-up. 1 RHA and 1 THA were revised due to pseudotumour formation, 2 THA were revised because of recurrent dislocations and 1 RHA for avascular necrosis. Overall an initial decrease in BMD was observed for both implants, stabilising after 2 years. 5 years after RHA a BMD change of +1% in upper cranial, -4% ( p < 0.01) in cranial, -8% ( p < 0.01) in craniomedial, -7% ( p < 0.01) in medial and +4% in caudal ROI compared to baseline values was seen. 5 years after THA a BMD change of -3% ( p = 0.01), -13% ( p < 0.01), -21% ( p < 0.01), -11% ( p < 0.01) and -2% for each respective ROI. The observed BMD decrease in different regions was structurally favouring the RHA-cup, with significantly higher levels in the cranial and craniomedial ROI. Acetabular BMD is better preserved behind a rigid press-fit convex cup in RHA compared to a titanium threaded cup in conventional THA in the cranial and craniomedial ROI. Despite of a theoretical higher stress-shielding behind the stiff acetabular component in RHA, compared to the more elastic threaded titanium THA-cup, bone depletion behind the RHA component does not seem to be of major concern. EudraCT (2006-005610-12).