This thesis describes metabolic changes occurring in chromaffin cells when secreting catecholamine (principally adrenaline), and the factors involved in maintaining the rate of secretion. In perfused pig adrenal glands, <sup>31</sup/>p nuclear magnetic resonance showed that nucleotide stored with catecholamine in the secretory vesicles (chromaffin granules) of the chromaffin cell was distinguishable from cytoplasmic nucleotide. Intragranular pH was 5.52 ± 0.15 (± SD, n=8) in ischaemic glands and rose (+ 0.22 ± 0.16 (± SD, n=6)) on recovery of cytoplasmic ATP during perfusion. This suggests that catecholamine accumulation by the granules is not driven by an ATP-generated pH gradient in intact tissue, as cytoplasmic ATP did not reduce intragranular pH. Perfused cortex-free ox adrenal medulla consumed 0.51 ± 0.19 (± SD, n=8) μmole 0₂/min/g wet weight after 210-230 minutes of perfusion, and this rose 30% during 4 minute O.lmM acetylcholine stimulations. This enhancement correlated with secretion but depended on the mode of stimulation, indicating that ATP consumption in secretion itself was an inadequate explanation. The proton-translocating Mg-ATPase of the chromaffin granule may hydrolyse ATP at its uncoupled rate on entering the plasma membrane during secretion by exocytosis. 1.4 ± 0.9 (± SD, n=12) moles of catecholamine were secreted per mole of enhanced oxygen consumption over 16 minutes. From this ratio, the oxygen consumption enhancement is shown to be much larger than that predicted from uncoupled proton pumping. Ouabain-sensitive oxygen consumption rose from < 6% to 18 ± 8% (± SD, n=4) during prolonged acetylcholine stimulation in the absence of calcium, suggesting that Na,K-ATPase was not responsible for all of the oxygen consumption enhancement. On continuous stimulation, secretion showed a biphasic decline in both pig and ox. A decline was also observed on intermittent stimulation. Cell death, potential-sensitive calcium gating and acetylcholine receptor desensitisation were only minor contributors. Little recovery occurred on resting the tissue for 2-3 hours between stimulations. The results are explained in terms of depletion of a pool of chromaffin granules adjacent to the plasma membrane.