The shafts of ulnae from 110 g male rats were cultured, and after a period of 5 h preincubation one of each pair of bones was either loaded cyclically (500 g, 1 Hz, 8 minutes) to produce physiologic strains (-1300 mu epsilon) or treated with exogenous prostacyclin (PGI2) or prostaglandin E2 (10(-6) M, 8 minutes) in the presence or absence of 17 beta-estradiol (10(-8) M). PGI2, PGE2, and loading stimulated almost immediate increases in glucose 6-phosphate dehydrogenase (G6PD) activity in osteocytes and osteoblasts. This increase was uniform throughout the section with exogenous PGs in the medium but was related to local strain magnitude in loading. Elevated G6PD levels in response to loading and PGI2 persisted for 18 h, by which time, ALP activity in surface osteoblasts was elevated and [3H]proline incorporation into collagen increased. PGE2 produced similar immediate and sustained increases in G6PD activity and [3H]proline incorporation after 18 h but no change in ALP activity. Bones cultured for 18 h with 17 beta-estradiol increased their [3H]proline incorporation, as did those loaded, and treated with PGI2 and PGE2. Loading and PGI2 but not PGE2 produced similar proportional increases in [3H]proline incorporation above the increased baseline of estradiol alone. These results suggest that estrogen and loading together produce a greater osteogenic response than either separately. If so, estrogen withdrawal would result in a rapid fall in bone mass to establish a new equilibrium appropriate to the reduced effectiveness of the loading-related stimulus. Such a fall in bone mass is a characteristic feature of estrogen withdrawal at the menopause.