Mild primary hyperparathyroidism (PHPT) is a common disease especially in middle-aged and elderly women. The diagnosis is frequently made incidentally and treatment strategies are widely discussed. To study the effect of parathyroidectomy (PTX) compared with observation (OBS) on biochemistry, safety, bone mineral density (BMD), and new fractures. Prospective, randomized controlled study (SIPH study), with a 5-year follow-up. The study was conducted at multicenter, tertiary referral centers. Of 191 randomized patients with mild PHPT, biochemical data were available for 145 patients after 5 years, with a mean age at inclusion of 62.8 years (OBS group, 9 males) and 62.1 years (PTX group, 10 males). Parathyroidectomy vs observation. Biochemistry, BMD, and new radiographic vertebral fractures. Serum-calcium and PTH-levels normalized after surgery and did not deteriorate by observation. BMD Z-scores were normal at inclusion in the lumbar spine (LS) and femoral neck (FN). For LS, BMD Z-scores were stable for 5 years with observation, but decreased in FN (P < .02). After surgery, BMD Z-scores increased significantly in both compartments (P < .02 for both), with a highly significant treatment effect of surgery compared to observation (P < .001). During follow-up, five new clinically unrecognized vertebral fractures were found in 5 females, all in the OBS group (P = .058). Even though new vertebral fractures occurred only in the observation group, the frequency was not significantly different from the surgery group. Longer follow-up is needed before firm conclusions can be drawn about the long-term safety of observation, as opposed to surgery.