The purpose of this study was to investigate the performances on speaking assessments in a first semester Spanish course to determine if there are differences among four sections based on required speaking tasks and mode of class delivery. Two of the classes were conducted online in a distance education delivery mode and two of the classes were conducted in the classroom. One section of delivery mode was required to fulfill speaking tasks throughout the semester while the alternate sections were offered the same speaking tasks as optional activities. Furthermore, the study investigated testing items repeated prompts versus new prompts to determine if there were differences in performance. Finally, the study investigated student performance confidence and beliefs about course materials. The data were collected through quantitative methods (oral rubric scores) and qualitative methods (open-ended questions). A total of 64 students participated in the study. Eight students completed the distance education course with optional speaking tasks. Thirteen students completed the distance education course with required speaking. Eighteen students completed the classroom course with optional speaking, and 24 students completed the classroom course with required speaking. In general, all classes made improvement between the first assessment and the second assessment. However, the mean performance scores for three of the four speaking prompts did not show significant difference among the four classes. In addition, there were no significant differences in mean scores of performances at the introduction of the two new speaking tasks on the final assessment. Only with one speaking repeated measure was there a significant difference. The significant difference was between the classroom with required speaking section and the distance education with optional speaking section. With an analysis of the Target Language Use, it was determined that the differences were due to lexical complexity and morphological complexity. In general, the self-report of confidence levels on the final oral assessment showed no significant differences among the four classes. There was a significant drop in confidence on one of the two new assessment prompts for all classes. Student comments indicated that there were frustrations with cognitive overload and dealing with ambiguity. / text