Computer-based assessment (CBA) has practical and economical benefits in that it allows testing of a large student cohort with the facility of automated marking of responses. In 2001, we incorporated CBA as a means of promoting and measuring student understanding of course content. While anecdotal information suggested that student learning was enhanced and results showed that learning had occurred, the learning processes promoted by this technology needed explanation. We have tried to do this. In 2002, over 300 students in their second year of teacher education participated in formative and summative CBA while undertaking an English course. Data were collected to track learning pathways and effects on learning outcomes. Surveys, monitoring of student use of technology, and interviews provided data sources. Of particular interest were the ways in which students used on-line discussion group to engage in learning.