As part of a study of the effects of lithium carbonate on neurochemical function in man, platelet and lymphocyte adenylate cyclase activity and lymphocyte beta-adrenergic receptor binding characteristics were determined before and after 2 weeks of lithium treatment in 10 normal volunteers. Lithium had differential effects on platelet and lymphocyte adenylate cyclase activity. In platelets, basal and stimulated (guanyl imidodiphosphate [Gpp[NH]p] or cesium fluoride) adenylate cyclase activity was significantly augmented by lithium treatment. By contrast, in lymphocytes, Gpp(NH)p- and cesium fluoride-stimulated adenylate cyclase activity was unaffected, while basal activity was decreased modestly after lithium. These results are consistent with preclinical studies that suggest that lithium's effects on adenylate cyclase activity are specific with respect to tissue and brain region and that lithium may interfere with guanine nucleotide binding (G) protein function. Lithium treatment significantly increased the ratio of low- to high-affinity dissociation constants for agonist displacement of antagonist binding to lymphocyte beta-adrenergic receptors (thought to reflect coupling between the beta-adrenergic receptor and stimulatory G protein). Lithium had significant effects on measures associated with signal transduction that might be contrasted to its more subtle effects on neuronal function (norepinephrine release) and neuroendocrine systems (responses to serotoninergic challenge) in these same subjects (reported in a companion article). Lithium's primary site of action may be on signal transduction mechanisms. These effects subsequently may be manifested in changes in neurotransmitter function that may be important to lithium's mood-stabilizing actions.