Pain researchers who had accepted Sokolov(1963)'s match-mismatch model in order to explain pain behaviors differentiated overprediction from underprediction of pain intensity. Especially after underpredicting pain intensity, the subjects overpredicted subsequent pain intensity, perceived pain to be more intensive, and their perceived pain intensity was dishabituated. And they tended to show withdrawal from pain stimulus. Moreover, it· is likely that the more they confide in thier prediction of pain intensity, the more they experience negative effect when they overpredicted pain intensity. For instance, when the pain is less controllable than expected, expectations of next pain control will probably be lowered, thus people will experience pain to be more aversive. On the basis of these findings, this study hypothesized that the underexpectation of pain control and the overprediction of pain intensity had negative effects on pain responses. The subjects, who were 61 undergraduates, received electrical shock in order to overpredict pain intensity and to underexpect pain control. The results showed that after underpredicting the pain intensity, they tended to overpredict pain intensity. Their subjective pain intensity and GSR were shown to be dishabituated. But there were no differences between the earlier overprediction and the later overprediction groups. And after subject overexpected pain control, their expectations for next pain control will be probably lowered. But after overpredictions, perceptions of pain intensity and GSR remain constant. These findings suggest that the match-mismatch model efficiently explained the perception of control of pain as well as prediction of pain intensity. In other words, overprediction of pain intensity and underexpectation of pain control have negative effects on pain responses. Also, these findings suggested that patients who underpredict pain intensity and overexpect pain control will experience more intense pain and produce accompanying pain behaviors in clinical settings.