Experiments were performed to investigate the nature of alpha- and beta-adrenoceptors in blood vessels supplying the posterior capsule of the chronically inflamed rabbit knee joint, and results were compared to the finding from previous experiments on the normal and acutely inflamed joint to assess any alteration which may occur in the adrenoceptor profile due to the chronic inflammation process. Electrical stimulation of the posterior articular nerve resulted in vasoconstriction that was completely blocked by phentolamine. This constrictor response was almost equally inhibited by prazosin and yohimbine. The dose-response curves to close intraarterial injection of alpha-adrenoceptor agonists showed a rank-order potency of adrenaline = clonidine = phenylephrine. The adrenaline dose-response curve was shifted to the right by administration of alpha-antagonists with a rank-order potency of phentolamine = prazosin = yohimbine. At this stage of experiments, there was an equal response of alpha(1)- and alpha(2)- adrenoceptors in blood vessels of the chronically inflamed rabbit knee joint. In another group of animals, the neurally mediated vasodilatation, which appeared after administration of phentolamine, was completely blocked by propranolol and was reduced significantly by ICI118551, but the effect of atenolol was not significant. The dose-response curve to close intraarterial injection of beta-adrenoceptor agonists showed a rank-order potency of: isoprenaline > salbutamol > dobutamine. The isoprenaline dose-response curve was shifted to the right by administration of beta-antagonists with rank-order potency of propranolol > ICI118551 > atenolol. These experiments showed a greater beta(2)-adrenoceptor response than beta(1)-adrenoceptor response in chronically inflamed rabbit-knee-joint blood vessels. Overall, compared to previous experiments on normal joint in which alpha(2)- and beta(1)-adrenoceptor responses predominated, and in acutely inflamed joint in which an equal alpha(1)/alpha(2) and beta(1)/beta(2) response was shown, in chronically inflamed joint the sympathetic constriction response was returned toward normal. No more alpha-adrenoceptor shift had happened, and the shift of beta(1) to beta(2) response continued.