We compared the somatomedin (SM) activity of perfusates of livers from normal and hypophysectomized (hypox) rats with and without the addition of growth hormone (GH) to the perfusions. Livers were perfused for a total of 3 hours, with a change of Waymouth's medium MB 752/l every 30 min, in a continously recirculating system. SM activity was estimated by the stimulation of the incorporation of sulfate by hypox rat costal cartilage incubated in vitro. Endogenous SM activity (no addition of GH) in the perfusates of the livers of normal rats was initially 136 +/- 17% above buffer levels (mean +/- SE) and fell gradually to near-buffer levels over the first 2 1/2 hours. Endogenous SM activity in the perfusates of livers from hypox rats was initially 72 +/- 10% above buffer levels, significantly lower than that of the normal livers, and reached buffer levels with the fifth medium change. Pretreatment of hypox rats with GH in vivo increased their initial liver perfusate SM activity to 134 +/- 25%, comparable to that of normal livers and significantly greater than that of livers of untreated hypox rats. The addition of GH (25 mug/ml) to the perfusion of both normal and hypox rat livers led to significant increases in perfusate SM activity compared with that of control perfusates at the same time periods. The increase in SM activity was not significant until after 0-30 minutes with hypox livers, and after 30-90 minutes with normal livers. Although it has previously been reported that the direct addition of GH to livers perfusion systems led to an increase in perfusate SM activity, we have now established that in the absence of added GH, perfusates of normal rat livers contain significant SM activity, and that perfusates of normal rat livers contain more SM activity than perfusates of hypox rat livers. In addition, the observation that GH pretreatment increases the SM activity of perfusates of hypox rat livers constitutes the first evidence that SM activity of any body tissue may be altered by administration of GH in vivo.