Intermittent parathyroid hormone (PTH) therapy increases bone mass. The purpose of this study was to determine if analogs of human PTH(1-34) (hPTH[1-34]), which differ from the native sequence in their receptor-activating properties, could promote bone formation in an ovariectomized (OVX) osteopenic rat model. We synthesized two hPTH(1-34) analogs with single substitutions for serine in the 3-position that in vitro are partial agonists in kidney. In the renal cell line OK, maximal cyclic adenosine monophosphate (cAMP) activation by [His(3)]hPTH(134) was 50%, and maximal cAMP activation by [Leu(3)]hPTH(1-34) was 20% of that produced by hPTH(1-34). Both analogs were full agonists in UMR-106 rat osteosarcoma cells and other bone-derived systems, but both had reduced potency compared with hPTh(1-34). Six-month-old retired breeder Sprague-Dawley rats were ovariectomized, and five animals underwent sham operation. On day 56 post-OVX, five sham-operated and five pre-PTH treatment OVX animals were sacrificed, and the remaining animals were randomized into 10 groups of six animals each. All other animals were injected with one of the hPTH analogs or hPTH(1-34) at 0, 4, 40, or 400 mu g/kg of body weight (BW)/day and were killed on day 84. Histomorphometry of the proximal tibia metaphysis and biochemical markers of bone turnover (osteocalcin and pyridinoline cross-links) were the primary endpoints. The cancellous bone volume was significantly lower at day 56 post-OVX (pretreatment) and at day 84 post-OVX (post-vehicle treatment) than at baseline. None of the compounds significantly increased the cancellous bone volume. Trabecular number declined after OVX and did not change with hPTH treatment. In contrast, the trabecular thickness declined after OVX but was higher after treatment with 40 mu g/kg of BW/day or 400 mu g/kg of BW/day of hPTH(1-34). In OVX rats, the mineralizing surface was higher than baseline at day 56 and fell toward control levels by day 84. All three peptides produced marked dose-related increases in the mineralizing surface and bone formation rates, but the two analogs were less potent than hPTH(1-34). Likewise, all peptides produced significant dose-related increases in the serum osteocalcin level. The osteoclast surface was not affected by OVX but was decreased with medium and high doses of hPTH(1-34). Pyridinoline cross-link excretion was not significantly affected by treatment with hPTH(1-34) but responded with a dose-dependent decrease to treatment with [His3]hPTH(1-34). These data suggest that bone selective analogs of hPTH(1-34) maintain the ability to induce bone formation but are less potent than hPTH(1-34).