Shock intensity (.3 or 1.6 ma.) was combined faotorially with ITI (15, 30, or 45 sec.) in a shuttle-box avoidance task (n -20/group). Following 65 avoidance trials, rats were allowed, in the absence of both shock and the CS, to jump from one compartment of the shuttle box to an adjacent safe box. The 1.6-ma. groups made significantly fewer avoidance responses than the ,3-ma. groups but jumped into the safe box with significantly shorter latencies. The latter finding is contrary to ari interpretation of the inverse relationship between shuttle-box avoidance learning and shock intensity which is based on freezing responses but supports a reinforcement interpretation which emphasizes the role of fear of situational cues.That there is an inverse relationship between strength of shock and avoidance learning by rats in a shuttle box is well documented (e.g.,