The COVID-19 crisis increased unemployment all over the World, with significant regional heterogeneity. This paper intends to analyze this territorial heterogeneity for the Portuguese case and investigate which regional factors complement personal and job characteristics in explaining individual vulnerability to COVID-19 unemployment. By considering personal, job and regional dimensions, we extended the literature and provided a more comprehensive understanding of this new phenomenon in the immediate and medium-term. Furthermore, this knowledge is essential to support policy suggestions for quick and effective action in preventing job losses in the current and future crises. Detailed information on all individuals that lost their jobs in Portugal 1 year after (and before) the COVID-19 outset was used to estimate three logit models that compare the odds of losing a job after and during the pandemic. Significant territorial heterogeneity of the COVID-19 impact on unemployment is obtained. Along with personal and job characteristics, we conclude that regional characteristics are essential for explaining individual vulnerabilities. In particular, workers are more prompted to lose their jobs if they live in regions with higher population densities, lower pre-crisis unemployment, and more dependable international flow. Conversely, individual and regional human capital investment contributes to protecting employment, revealing the existence of external effects.