A method of teaching English as a second language (ESL) involving open sentence combining practice with group discussion was compared with a more traditional teaching method. Students were largely college-bound and attended the University of Arizona's ESL Center. Sentence combining was analyzed for a paragraph-length exercise on a low-level scientific topic and a free-writing sample on a daily conversation topic, which were produced in the first and last weeks of an 8-week session. For 7 weeks, the traditional instruction (control) group completed text exercises and some compositions and identified and corrected errors as a class. During this same time span, the experimental group completed a free-writing assignment focusing on some particular structure or an open-writing exercise in sentence combining using a variety of structures. They were also given an essay assignment designed to elicit a given structure. It was found that the type of sentence combining practiced by the experimental group resulted in increased syntactic maturity and a decrease of error. The control group sustained losses, particularly in the error-free measures as they attempted to use the complex structures that they had studied but not mastered. Attention is directed to types of errors, preliminary findings of a 16-week study, and a hypothesis concerning the relationship between increased syntactic maturity scores and improvement in reading comprehension. (SW)