The rapid pace of technological development makes work-related training an increasingly important source of competences for employees. It seems that new competences may have been crucial during the COVID-19 pandemic, as they provided employees with the knowledge and skills needed to work in a hybrid or remote form. Therefore, from a theoretical point of view, one would expect that work-related training contributed to a decrease in the risk of job loss at that time. Using individual data from the Polish Labour Force Survey (LFS) for the years 2018–2020, we estimated a probit model of outflows from employment to see whether participation in work-related training reduced the probability of outflows from employment into unemployment. However, the results do not confirm the existence of such a relationship either before (2018–2019) or during (2020) the pandemic.