The Fear Survey Schedule, the State-Trait Anxiety Inventory, and a situational fear inventory were administered to 65 junior high school students in Beirut shortly before the Israeli invasion of 1982 and again 8 wks after the disengagement of forces. 48 Ss were evacuees and 17 were nonevacuees. There were no significant differences between either the pre- or postinvasion scores of the evacuees and nonevacuees. Aggregate anxiety estimates of both groups with respect to war-related stimuli were significantly lower at postassessment. Findings are viewed as incompatible with the classical conditioning theory of fear acquisition, in that the direct exposure of nonevacuees to wartime stress did not result in greater fear or anxiety, but as consistent with a self-efficacy conceptualization. Results also agree with previous reports concerning the effects of exposure to war-related stress.