Research with the Communication Attitude Test and a Dutch translation (C.A.T.-D) has shown that, from age 6 to 14, children who stutter show significantly more malattitudes toward speech than nonstutterers. Investigations also have shown that the test procedure has inter-item reliability and items that correlate significantly with the total score. Yet to be determined, however, is the C.A.T.'s test-retest reliability and its sensitivity to attitudinal change. To determine this, forty-four Dutch speaking Belgian grade-schoolers responded to the questionnaire's items on three separate occasions. The test was re-administered again 1 and 12 weeks later. The reliability coefficients for the subjects' test scores following a hiatus of 1, 11 and 12 weeks was +.83, +.81 and +.76, respectively. These statistically significant correlations show that the speech-associated attitudes of the stuttering children were positionally stable. Over a 1 week period their attitudes also evidenced absolute stability. However, with respect to the test's sensitivity, the significant absolute mean differences following an 11 and 12 week time period suggest that the test is capable of reflecting a change in speech-associated attitudes.