It has been suggested that the magnitude and form of hypoalgesia elicited by an aversive event can be predicted from its coulometric product (Intensity X Duration). According to this hypothesis, small products elicit opioid hypoalgesia, and huge products elicit nonopioid hypoalgesia. This suggests that increasing the duration of an aversive event should heighten the nonopioid hypoalgesia. Contrary to this prediction, in Experiment 1 I found that increasing tbe duration of a mild shock attenuated the nonopioid hypoalgesia. In Experiment 2 I tested another implication of the coulometric hypothesis, namely, tbat mild shocks that have the same coulometric product should elicit equivalent hypoalgesia. The results did not support this prediction. We discuss how tbese findings are consistent with an alternative theory, the "working memory hypothesis." According to this theory, the representation of an aversive event in working memory elicits hypoalgesia. In Experiment 3 a novel prediction of this theory was tested, namely, tbat displacing the representation of intense shock from working memory, by following the intense shock with a weak shock "distractor," sbould attenuate hypoalgesia. The results support this prediction. I conclude by discussing the relation of this work to other findings in the analgesia literature.