1. Renin-containing granules were isolated, characterized, and used to gain insight into a possible chemiosmotic mechanism of renin secretion. 2. Renin granules were obtained by a modification of the sucrose gradient method, which yielded a 67-fold purification of renin granules as assessed by marker enzymes, or a modification of the Percoll gradient, which yielded a 230-fold enrichment of renin granules. 3. Granular renin content was increased by chronic sodium deprivation and hypophysectomy. 4. Renin release from granules was inversely related to osmotic strength (150-900 mosmol l-1). pH had a biphasic effect on renin release, with greater stimulation at both acidic (pH 5) and alkaline (pH 8 and 9) pH. The pH effect was dependent on Cl-; raising Cl- stimulated release. This effect was abolished by-oligomycin and N,N'-dicyclohexylcarbodiimide (DCCD) at pH 5, but not at pH 8; the effect was enhanced by NH4+. 5. Either valinomycin or carbonyl cyanide m-chlorophenylhydrazone (CCCP) alone was without effect; but in combination they caused a potent stimulation at all pHs. Nigericin stimulated renin release at all pHs, but its effect required K+. 6. Raising K+ stimulated renin release from granules, whereas raising Na+ was without effect. Lowering Ca2+ below 10(-6) M significantly stimulated renin release. 7. Taken together, the evidence is consistent with the chemiosmotic hypothesis for the control of renin release from granules and may have some implications for the regulation of renin secretion from juxtaglomerular cells.