The first aim of the study was to examine what sort of effect prior exposure to conflict has on succeeding performance in a weak conflict situation, Dynamogenic and competing response hypotheses were used to predict the performance of subjects in the subsequent conflict situation. The dynamogenic view asserts that conflict generates a drive, thus the effect of conflict training is to facilitate the subsequent performance, reducing the reaction time in post-conflict situation. According to the competing response view, conflict training leads to learning of specific conflict responses, which are generalized to postconflict performance, with the resulting inhibition of performance, or the increase of reaction time. Three groups of normal subjects were run under three ssuccesive phases of experiment. First, all the groups were exposed to a low conflict situation. Following this, each of the groups was given different degree of conflict training, such as high, moderate, or low. Finally, all the groups were again exposed to the original low conflict situation. The comparison of the conflict effect between normal and abnormal subjects, which was the second aim of this study, was performed by running groups of neurotic and schizophrenic subjects under the same conditions as above. The modified Stroops's Color-Word Test cards were used for inducing different degrees of conflict, designated respectively as the high conflict card (HC), the moderate conflict card (MC) and the low conflict card (LC). On these conflict cards the color names were printed in colors. The HC card had no word printed in the corresponding color, while the LC card had all the words printed in the corresponding colors, and the MC card had printed words not corresponding to the colors, but in colors suggesting the same hind of color. The subject was instructed to identify the color of each stimulus irrespective of the color designation. The author hypothesized that the conflict induced by the experiment was a type of approach-approach conflict between two response tendencies; the approach toward reading the word which was motivated by cultural and verbal habit and the other approach toward the color itself which was motivated by the instructions. The reaction time (1/RT×1000) for each conflict card was used as a measure of conflict. The findings are shown in Tables 1 to 6 and Figures 1 to 3. The results indicate that the effect of strong and moderate conflict training led to significant retardation of the reaction time in the postconflict performance for all groups. This fact supports the competing response view, but also has implications for the dynamogenic view.