Purpose: The aim of this study was to investigate the correlation between (a) the work-related feelings and behaviours and (b) workaholism and work engagement. The correlation model was based on the Affective Events Theory, proposed by Weiss and Cropanzano. So far, workaholism and work engagement have not been investigated in terms of this concept. Design/methodology/approach: The study was based on a sample of 292 workers representing various occupations and industries. Four diagnostic questionnaires were used, namely DUWAS (for workaholism, or work addiction), UWES (work engagement), WORAF (for feelings) and WORAB (work-related behaviour). In order to verify the hypotheses, a path analysis was performed and twelve (12) models were designed. The indices of fit were good and/or acceptable for seven (7) of these models. Findings: The majority of the hypotheses were confirmed fully or partially. A relatively strong direct correlation was found between emotions and work-related behaviour. The mediation effect of workaholism and work engagement was rather weak. The study confirmed the positive nature of work engagement. However, workaholism was marked by ambiguity in this respect. Research limitations/implications: The main limitation of the study is the use of questionnaires. In the future, it would be advisable to use also other research methods and techniques to give credibility to the results. Another weakness of this study is that the sample is not representative. While respondents represented a variety of industries, occupations and positions, many others, particularly lower-skilled jobs (such as construction workers, shop assistants or warehouse operatives) were not covered. Originality/value: This study is the first one to have investigated workaholism, work engagement and work-related behaviours and feelings at the same time, which makes it, to some extent, a pioneering piece of work.