The effects of a new anxiolytic, (2-(7-chloro-1,8-naphthyridin-2-yl)-3- [(1,4-dioxa-8-azaspiro[4.5]dec-8-yl)-carbonylmethyl] isoindolin-1-one (DN-2327), on the execution of step-through passive avoidance and delayed spontaneous alternation tasks were assessed and compared with those of diazepam (DZP) and buspirone. DN-2327 and buspirone (both 10 and 20 mg/kg, PO) impaired performance in the 48-h passive avoidance recall test when given prior to the test session, but not when given before the training trial. DZP impaired the performance at doses of more than 5 and more than 10 mg/kg PO when given prior to the test session and when given before the training trial, respectively. The action of DZP (10 mg/kg PO) when given before the training trial was antagonized by flumazenil (20 mg/kg, IP) and tended to be antagonized by DN-2327 (10 and 30 mg/kg, PO), but was not affected by buspirone. No evidence for possible amnesic effects of DN-2327 or buspirone on working memory was found in the delayed spontaneous alternation task, but DZP (3 and 10 mg/kg, PO) caused significant impairment of working memory. Electroshock sensitivities detected by flinch, jump, and vocalization thresholds were not influenced significantly by DN-2327 (30 and 100 mg/kg, PO), DZP (10 and 30 mg/kg, PO) or buspirone (30 and 100 mg/kg, PO).(ABSTRACT TRUNCATED AT 250 WORDS)