Both ulnas of skeletally mature roosters (Gallus domesticus) were deprived of functional load bearing by proximal and distal submetaphyseal osteotomies. Twenty-four hours later the animals were injected with 1.5 mCi of3H-uridine and the ulna on one side was subjected to a single period of a cyclical load engendering physiological strain levels at 1 Hz for 6 min. Twenty-fours after loading the animals were killed. Autoradiographic examination of comparable regions of cortex in sections from the bones' midshafts showed that in the loaded bones, 72±2.7% of osteocytes were labeled compared with 12±3.5% in the corresponding areas of their contralateral nonloaded pair (P<0.001). The number of grains per labeled osteocyte was also higher in the loaded side (6±0.5 compared with 4±0.5,P<0.01). There was no obvious correlation between the longitudinal strain distribution during artificial loading and the distribution of labeled osteocytes throughout the bone cross-section. However, previous long-term experiments using a similar loading preparation had consistently shown the site of most periosteal new bone formation to also not be directly related to the local strain magnitude. Perhaps it is significant that the greatest percentage of labeled cells were found in the cortex where the long-term experiments had shown most new bone formation to subsequently occur.