Performance on working memory tasks, a measure of prefrontal cortical function, is impaired by exposure to mild stress as well as the anxiogenic drug, FG7142. Previous studies have shown that like stress, FG7142 increases catecholamine release in the prefrontal cortex (PFC) and that high levels of dopamine (DA) D(1) and norepinephrine (NE) alpha-1 receptor stimulation underlie the FG7142-induced cognitive impairment. Both the FG7142-induced DA turnover and working memory deficit can be blocked by pretreatment with the nonselective NE alpha-2/imidazoline I1 receptor agonist, clonidine. The present study examined the alpha-2 adrenoceptor subtype underlying this reversal in FG7142-induced working memory deficits by comparing the efficacy of clonidine with the more selective alpha-2A adrenoceptor agonist, guanfacine. The anxiogenic drug, FG7142 (0, 10, 20, or 30 mg/kg), dose-dependently impaired delayed alternation performance. Clonidine pretreatment (0.1 mg/kg, 30 min prior to FG7142) partially reversed the FG7142-induced impairment while guanfacine pretreatment (0.11 mg/kg) completely blocked the FG7142-induced impairment. Neither clonidine nor guanfacine had any effect on performance when administered alone. This study suggests that stimulation of the NE alpha-2A receptor subtype is sufficient to ameliorate the cognitive deficit induced by FG7142. Clonidine's sedative and hypotensive side effects limit its therapeutic usefulness; however, selective alpha-2A receptor agonists may be effective in treating prefrontal cognitive deficits in stress-related neuropsychiatric disorders with fewer side effects.