Two control classes and two experimental classes of first-semester college freshmen participated in this study, which examined the effect of expressive writing on attitudes toward writing in general. Both groups were taught the same content material and wrote the same number of papers. In addition, the experimental group did expressive writing in nongraded journals. The results of the study indicate that the attitudes of both groups improved overall, with the exception of attitudes toward English, which became slightly worse. The attitudes of the experimental students improved significantly with regard to writing; control students improved significantly with regard to self-concept and belief in expressive ability. Upon further analysis, it was found that the self-reliance portion of the measurement instruments made this difference in the control group's attitudes. In evaluations of writing, the trend was for the experimental group's papers to receive higher ratings than the control group's papers, with respect to both overall writing ability and specific writing attributes. Moreover, the experimental group's papers were found to have improved significantly with regard to concreteness in writing. (Author/KS)