Osmotic pumps were evaluated for 7-d delivery of growth hormone-releasing factor (GRF). In Exp. 1, 12 steers weighing 253 kg received hGRF(1-29)NH2 in H2O at rates of 0, 3, 30 and 300 pmol.h-1.kg-1. Pumps were implanted s.c. on d 0 and removed at 1200 on d 7. Blood samples were drawn at 20-min intervals from 0800 to 1200 on d -1, 1, 3, 5, 7 and 9. Growth hormone levels were not altered by GRF treatment (P greater than .05). Solubility and volume limitations render hGRF(1-29)NH2 delivery via osmotic pumps problematical. Flow rate and duration of release of dimethyl sulfoxide (DMSO):H2) (1:1) from osmotic pumps incubated in vivo and in vitro were found to be consistent with manufacturer's specifications. Two hGRF(1-29) analogues, Ro23-7863 and 4SG-29, were dissolved in DMSO:H2O. In Exp. 2, six 222-kg steers had pumps implanted and blood samples were taken as in Exp. 1. Three steers received each analogue at a rate of 300 pmol.h-1.kg-1. Analogues had similar GH-releasing ability and GH levels differed (P less than 0.001) among days, being approximately fourfold higher on d 3, 5 and 7 than on d -1, 1 and 9. Residual analogue solutions retained full bioactivity after 7-d implantation, and in vitro biopotencies of Ro23-7863 and 4SG-29 were similar (Exp. 3). In Exp. 4, 15 wethers (means = 31.3 kg) received osmotic pumps delivering 0, 3, 15, 75 and 300 pmol.h-1.kg-1 Ro23-7863 in DMSO:H2O for 7 d. Lambs were bled at 0800 and 1400 from d -1 to 8. The latter two doses increased (P less than .01) mean GH levels 2.7- and 4.3-fold over those in control animals during the treatment period. Results demonstrate that increased GH secretion can be elicited in steers and wethers for 1 wk by continuous s.c. infusion of GRF analogues utilizing osmotic pumps.