With the COVID-19 pandemic, the physical activity (PA) levels of university students declined as a result of confinement. The aim of the study was to analyse the segmentation of university students according to physical self-concept ratings and to determine the differences between each cluster during the pandemic. The sample consisted of 492 students aged 18–31 years, 36.8% male and 63.2% female, who were administered the PSDQ-S and IPAQ instruments. The data collected were analysed with SPSS software, from which descriptive statistics, a cluster analysis from the PSDQ-S, was obtained. The IPAQ and socio-demographic variables were used to characterise the groups. Finally, a non-hierarchical K-means analysis was performed to establish the clusters. The results reported three different group profiles of students. Significant differences were found in all self-concept variables analysed, with the exception of some items. In relation to PA level, it could be established that the Positive Physical Self-Concept group had the highest PA level and was composed of 52.1% females and 34.4% males, showing a high physical self-concept, whereas, in the Medium Physical Self-Concept and Negative-Physical Self-Concept groups, females were predominant in number. They were also the least physically active groups and had a low physical self-concept.