Abstract Objective There has been increased attention in recent years to mental health, quality of life, stress and academic performance among university students, and the possible influence of learning styles. Brief reliable questionnaires are useful in large-scale multivariate research designs, such as the largely survey-based research on well-being and academic performance of university students. The objective of this study was to examine the psychometric properties of a briefer version of the 39-item Adelaide Diagnostic Learning Inventory. Results In two survey samples—medical and physiotherapy students—a 21-item version Adelaide Diagnostic Learning Inventory-Brief (ADLIB) was shown to have the same component structure as the parent instrument, and the component structure of the brief instrument was found to generalise across students of medicine and physiotherapy. Subscale reliability estimations were in the order of magnitude of the parent instrument. Subscale inter-correlations, inter-component congruence coefficients, and correlations between ADLIB subscale scores and several external measures provide support support for the construct and criterion validity of the instrument.