The primary purposes of this study were (1) to determine whether significant differences existed among second-grade experimental and control groups in the rated quality of their compositions and in their use of selected syntactic structures at the end of a one year experimental program in written composition, and (2) to determine the relationships between rated quality and syntactic complexity of compositions written by these second-grade children. As a secondary treatment of the data, the syntactic analysis and quality ratings were related to subject variables: sex, chronological age, reading achievement, IQ, and socioeconomic level. The population consisted of second-grade students. Statistical techniques used to treat the raw data were analysis of covariance and Pearson product moment correlations. Conclusions are: (1) Significant differences existed among experimental and control groups in the rated quality of their compositions. (2) Significant differences existed among experimental and control groups in number of words, number of T-units, subordinate clauses, verbal phrases, adverbials, prepositional phrases, and initial coordinating conjunctions. (3) There is a significant relationship between composition quality and several syntactic variables, and between IQ, and reading achievement. There is no significant relationship between subject variables and syntactic structures. (Author/DB)