Early postmenopausal women ( n = 160) were randomised to receive placebo or strontium ranelate (SR) 125 mg/day, 500 mg/day or 1 g/day for 2 years (40 participants per group). All participants received calcium 500 mg/day. The primary efficacy parameter was the percent variation in lumbar bone mineral density (BMD), measured using dual-energy X-ray absorptiometry. Secondary efficacy criteria included hip BMD and biochemical markers of bone turnover. At month 24, SR 1 g/day significantly increased lumbar BMD compared with placebo [mean (SD) +5.53% (5.12); p<0.001] for measured values and [mean (SD) +1.41% (5.33%); p<0.05] for values adjusted for bone strontium content. The annual increase for adjusted values was +0.66% compared with -0.5% with placebo, with an overall beneficial effect after 2 years of about 2.4% with SR 1 g/day relative to placebo. There were no other significant between-group differences in adjusted lumbar BMD. Femoral neck and total hip BMD were also significantly increased at month 24 with SR 1 g/day compared with placebo [mean (SD): +2.46% (4.78) and +3.21% (4.68), respectively; both p<0.001)]. SR 1 g/day significantly increased bone alkaline phosphatase at all time points ( p<0.05) compared with baseline and between-group analysis showed a significant increase, compared with placebo, at month 18 ( p = 0.048). No effect on markers of bone resorption was observed. SR was as well tolerated as placebo. The minimum does at which SR is effective in preventing bone loss in early postmenopausal non-osteoporotic women is therefore 1 g/day.