Objectives: This study aimed at investigating cross-sectional relationships between psychosocial characteristics of work and presenteeism in a sample of Belgian middle-aged workers. Material and Methods: Data were collected from 1372 male and 1611 female workers in the Belstress III study. Psychosocial characteristics assessed by the use of self-administered questionnaires were: job demands, job control, social support, efforts, rewards, bullying, home-to-work conflict and work-to-home conflict. Presenteeism was measured using a single item question, and it was defined as going to work despite illness at least 2 times in the preceding year. Logistic regression models were used to investigate the relationship between psychosocial characteristics and presenteeism, while adjusting for several socio-demographic, health-related variables and neuroticism. An additional analysis in a subgroup of workers with good self-rated health and low neuroticism was conducted. Results: The prevalence of presenteeism was 50.6%. Overall results, adjusted for major confounders, revealed that high job demands, high efforts, low support and low rewards were associated with presenteeism. Furthermore, a significant association could be observed for both bullying and work-to-home conflict in relation to presenteeism. The subgroup analysis on a selection of workers with good self-rated health and low neuroticism generally confirmed these results. Conclusions: Both job content related factors as well as work contextual psychosocial factors were significantly related to presenteeism. These results suggest that presenteeism is not purely driven by the health status of a worker, but that psychosocial work characteristics also play a role.