Purpose: This paper aims to investigate the relationship between employee training and bank risk to verify whether and to what extent an increase in employee training, as a soft component of total quality management (TQM), affects bank risk. Design/methodology/approach: The research adopts a panel regression, based on a unique dataset of a sample of Italian banks over the period 2011–2018, to test whether employee training affects bank risk, measured alternatively in terms of Z-score, a proxy of bank stability and non-performing loans (NPLs)/gross loans ratio as a proxy of credit risk. Findings: Research findings reveal that increasing employee training leads to growing bank stability. In contrast, credit risk is not affected by employee training. However, by investigating training heterogeneity, this study found that the increase in the number of managerial training hours, as a proxy for soft skills training, negatively impacts credit risk. Therefore, an increase in soft skills leads to a reduction in bank credit risk. Research limitations/implications: This study provides empirical evidence in support of the relationship between employee training and bank risk, which seems novel in the literature. From a managerial point of view, this study highlights the need for banks to pay attention to the skills, particularly soft skills, that banks' employees must possess to effectively manage bank risk and, more specifically, the core bank risk. Originality/value: Empirical evidence on the relationship between employee training, soft/hard skills and bank risk appears limited if not absent. Therefore, the findings provide insights for a more nuanced interpretation of variables that affect bank risk. [ABSTRACT FROM AUTHOR]