At-risk children in danger of experiencing failure in reading, face negative life consequences that may be alleviated by the inclusion of an attributional training program that would increase the possibilities of success. The present study investigated the causal attributions of students in a Success for All (SFA) school versus a traditional reading curriculum. A total of 197 children, 89 from the SFA experimental school, and 108 control children participated in this study. Data were collected through the posttest administration of the Causal Beliefs About Reading Questionnaire based on the attribution work by Ames (1978, 1981, 1984) and others. Pretest reading achievement (Woodcock, Durrell) and diagnostic measures (PPVT) from the broader study by Chambers et al. (1996) were used as covariates, since pretest causal data were not available. The correlations indicated some interesting patterns. Of significance, were the negative correlations between ability and task difficulty, and the positive correlations between ability and effort. While the ANCOVA results for the hypotheses were not significant, further ANOVA results employing ratings of children's perceptions of success in reading were significant. In particular, there was a 2-way interaction effect for the type of reading program and perceptions of success in reading on attributions to the difficulty of the task. These findings suggest important motivational and feedback techniques that should be taken into consideration by classroom teachers. The key is for teachers to use research-based teaching techniques in combination with effective attributional practices.