This longitudinal study compared the oral production of undergraduate EFL students in an online distance learning program and a classroom-based learning program, offered by the Federal University of Ceará (Brazil) for EFL teacher certification. The students from both programs completed oral tasks similar to the TOEFL iBT oral independent questions at the end of each term of study, from first to fourth semester. Their oral productions were analyzed as to fluency, accuracy, and complexity. For fluency, the produced words were divided by time spent completing the task, excluding time for pauses. Two accuracy measures were employed, a syntactic measure (the average of correct clauses per task), and a lexical measure (number of mistaken words divided by number of words produced). For complexity, two measurements were also used, a syntactic measure (number of clauses produced divided by number of AS-units), and a lexical measure (number of different words divided by number of words produced for each task). ANOVA of repeated measures was used to analyze fluency, accuracy, and complexity. The ANOVA assessed the effects of three variables: Group (classroom-based and distance), Time (semesters one through four), and Groups over Time. No significant differences in lexical accuracy and lexical complexity were found between groups, although the groups presented differences in syntactic accuracy, syntactic complexity and fluency. For the analysis of the variable Time, results obtained by all students were analyzed. No significant differences in lexical accuracy and lexical complexity were found for the tasks completed in each of the four semesters, although the completion of the tasks presented differences in syntactic accuracy, syntactic complexity and fluency. When comparing the results obtained by the groups over time, no significant differences were found in each of the parameters; in other words, both groups presented similar development in their oral production.
