the one hand and sets of intentions on the other, Fishbein and Ajzen(1974) asked subjects to indicate whether they would or would not perform each of 100 behaviors that had previously bee judged as indicating either a favorable or an unfavorable religious attitude. Taking this evaluation of the behavior into account, they computed an index by summing across responses to all 100 intentions. In ad- dition, subjects filled out five traditional verbal measures of attitude toward reli- giosity(a self-report scale, a semantic differential, and Likert, Guttman, and Thurstone scales). The correlations between these measures of attitude and the index based on the 100 intentions ranged from.60 to.75. In marked contrast, the average correlation between attitude and single intentions ranged from.16 to.20 for the different attitude scales. In conclusion, there appears to be no systematic relation between attitudes and intentions. Although attitudes tend to correlate highly with indices based on sets of intentions, the relation between attitude and single intentions is usually low and nonsignificant. To gain a better understanding of the relation between atti- tudes and intentions, it may be necessary to examine the nature of intentions in