The purpose of the present study was to examine the problems of cognitive interference effects related to intelligence. The modified Stroop Color Word Test (SCWT) and Intelligence Tests were administered to over 38 elementary school pupils for Experiment 1 and 45 mentally retarded pupils in junior high school for Experiment 2. The cognitive interference effects of Ss were inferred from their performance on the SCWT. The SCWT consists of three cards. On the card A four color names; red, blue, green and yellow are printed in black in Japanese. On the other hand card B has the same four colors. Card C is the interference card Which has no color names printed in the corresponding colors. Ss were instructed to read aloud as rapidly as possible the color names on card A and colors on card B and card C irrespective of color names. All Ss in Experiment 1 received a retest of the SCWT after one year. Reaction times and Error responses taken for each of the three cards were recorded as the measure of the SCWT. The main results were as follows: The reliabilities of the measures (RT) of the SCWT were determined by Spearmans Coefficient (rs) of rank correlation between the first and second administrations in Experiment 1. The RTs of the SCWT showed great developmental and practice effects, but the reliabilities of the RTs of the SCWT got to a highly significant level. The RT taken on the card C was differentially greater, than the RTs taken on the cards A and B. This holds both for normal and mentally retarded pupils analyzed separately. The intercorrelation coefficients of RTs of three cards showed high scores for normal pupils and for mentally retarded pupils. The rs between RTs of the SCWT and Standard Scores of Intelligence tests showed high negative scores in Experiment 1. The mean RT of Boarderline pupils was smaller than the mean RTs of Imbecile and Moron Pupils in Experiment 2 It was clear that individual differences in cognitive interference might be profitably related to intelligence.