Aim: To investigate the effect of femoral head size on blood metal-ion levels caused by taper corrosion in metal-on-polyethylene total hip arthroplasty, comparing 36- to 44-mm heads with 32-mm heads. Methods: In a randomised, controlled, single-blinded trial, 96 patients were allocated to receive either a 32-mm metal head or the largest possible metal head (36–44 mm) that could be accommodated in the thinnest available vitamin E, cross-linked polyethylene insert. Blood metal ion levels were collected at 1- and 2-year follow-ups. Results: At 1-year, metal-ion levels did not differ between the groups. The median (interquartile range) blood-ion levels for the 32-mm versus the 36- to 44-mm group were 0.11 µg/L (0.08–0.15) versus 0.12 µg/L (0.08–0.22), p = 0.546, for cobalt, 0.50 µg/L (0.50–0.59) versus 0.50 µg/L (0.50–1.20), p = 0.059, for chromium and 1.58 µg/L (1.38–2.05) versus 1.48 µg/L (1.14–1.87), p = 0.385, for titanium. At 2 years, there was no difference either and the corresponding values were 0.15 µg/L (0.12–0.24) versus 0.18 µg/L (0.12–0.28), p = 0.682 for cobalt, 0.50 µg/L (0.50–0.50) versus 0.50 µg/L (0.50–0.57), p = 0.554, for chromium and 1.54 µg/L (1.16–1.87) versus 1.42 µg/L (1.01–1.72), p = 0.207 for titanium. Conclusions: The use of the largest possible metal head (36–44 mm) compared to a 32-mm head in metal–on-polyethylene bearings does not appear to elevate blood metal-ion levels up to 2 years postoperatively. As taper corrosion is probably time-dependent, longer-term reports are needed to evaluate the association between large metal heads and blood metal ion levels. Trial registration: ClinicalTrials.gov (reg. ID NCT0231 6704).