A series of (+/-)-4,5-dihydro-4-phenyl-3H-1,3-benzodiazepines and (+/-)-4,5-dihydro-4-phenyl-1H-1,3-benzodiazepines was synthesized as part of a program to develop novel psychotropics. Of these compounds, (+/-)-4,5-dihydro-2,3-dimethyl-4-phenyl-3H-1,3-benzodiazepine (10a, HRP 543) emerged as a potential antidepressant. In in vivo mouse tests (inhibition of tetrabenazine-induced ptosis and potentiation of yohimbine toxicity) which are predictive of antidepressant-like activity, 10a is comparable to amitriptyline. The similarity is also maintained in vitro, as both 10a and amitriptyline inhibit norepinephrine and serotonin uptake into rat brain synaptosomes. No significant inhibition of rat brain monoamine oxidase A or B was found with 10a, nor did the compound potentiate tryptamine-induced seizures. On chronic administration, the number of cortical beta-adrenergic receptor sites was similarly reduced by 10a and desipramine. The anticholinergic activity of clinically useful antidepressants, such as amitriptyline, is a proposed cause of side effects which reduce patient compliance. In contrast to the tricyclics, 10a apparently lacks anticholinergic activity, as evidenced in vitro by negligible displacement of [3H]quinuclidinyl benzylate from rat brain muscarinic receptors and in vivo by insignificant antagonism of the cholinergic stimulation produced by physostigmine or oxotremorine. These data suggest that 10a may be clinically useful as a novel nontricyclic antidepressant which is devoid of anticholinergic side-effect liability. Further evaluation of 10a in nonrodent species is in progress.