Throughout much of the literature on plagiarism in higher education, there is an implicit assumption that students who understand plagiarism, who have high ethical views and declare not to engage in plagiaristic behaviour are able to recognise it and avoid it in practice. Challenging this supposition, this paper contrasts students self-reported data with their ability to recognise and avoid plagiarism in a proposed case scenario. A questionnaire was adapted from previous literature and administered to a sample of undergraduate first- and second-year students in an Irish university. Results show that self-reported measures are not a powerful predictor of the students ability to recognise the practical case as an academic breach, nor to avoid the breach through referencing. This suggests that students understanding and awareness of academic breaches would benefit from experiential learning and that higher education institutions should not merely rely on providing statements and definitions of academic misconduct. Also, the results highlight the potential unsuitability of using self-reported measures to study plagiarism, despite their widespread use.