The in vitro release of stored intracellular rat growth hormone (rGH) in response to several secretagogues suggests the functional division of rGH storage into at least two ‘compartments’ or ‘pools’. The first is an immediately releasable compartment whose response is brief. The second is a compartment which responds to more prolonged secretory demands. These observations are consistent with either a single, homogeneous population of somatotrophs, each of which exhibits functional compartmentalization of storage, or with heterogeneous populations of somatotrophs, each family of which provides one of the observed responses. We sought anatomical correlates of this functional compartmentalization using a perifusion/morphometric system which permitted examination of the first model while not excluding the second model. We selected for study an established perifusion protocol whose behavior was consistent and whose previous results suggested phases of both filling and emptying of the immediate release pool. Five parallel perifusions of pituitary fragments were run. The fifth perifusion was used to monitor rGH release and to confirm that the experiment had behaved in standard fashion. The first pituitary chamber was dismantled during basal perifusion to obtain tissue for microscopy, the next during exposure to 25 nM SRIF, the third during exposure to both 25 nM SRIF and 1 mM (Bu)2cAMP, and the fourth shortly after the rapid release which followed SRIF withdrawal. Somatotroph granulation was decreased by 54% in the presence of SRIF, and then increased by 45% with the addition of (Bu)2cAMP. The intracellular distribution of granules also fluctuated in relation to the stimulatory and inhibitory secretagogues. In addition, secondary lysosomes increased by 549% during SRIF-induced inhibition of rGH release. Cell volume was decreased by 37% in the presence of 25 nM SRIF; the addition of (Bu)2cAMP did not reverse this SRIF effect. We conclude that the in vitro release phenomenon we have termed the ‘immediate release pool’ has an anatomic correlate in juxtamembrane granules, that inhibition of rGH release is accompanied by a significant increase in secondary lysosome activity, that while heterogeneous populations of somatotrophs probably exist, individual cells possess a range of responses consistent with functional compartmentalization, and that rGH release involves not only granule movement between anatomically demonstrable compartments, but in addition, major and rapid changes in somatotroph volume.