Purpose: Self- and peer assessment has proved effective in promoting the development of teamwork and other professional skills in undergraduate students. However, in previous research approximately 30 percent of students reported that its use produced no perceived improvement in their teamwork experience. It was hypothesised that a significant number of these students were probably members of a team that would have functioned well without self- and peer assessment and hence the process did not improve their teamwork experience. This paper aims to report the testing of this hypothesis. Design/methodology/approach: The paper reviews some of the literature on self- and peer assessment, outlines the online self- and peer assessment tool SPARKPLUS, and analyses the results of a post-subject survey of students in a large multi-disciplinary engineering design subject. Findings: It was found that students who were neutral as to whether self- and peer assessment improved their teamwork experience cannot be assumed to be members of well-functioning teams. Originality/value: To increase the benefits for all students it is recommended that self- and peer assessment focuses on collaborative peer learning, not just assessment of team contributions. Furthermore, it is recommended that feedback sessions be focused on learning not just assessment outcomes and graduate attribute development should be recorded and tracked by linking development to categories required for professional accreditation. (Contains 4 figures and 1 table.)