During the Intervention Project on Absence and Well-being (IPAW), a 5-year project conducted in Denmark, we investigated psychosocial work environment factors as independent predictors of the number of absence days per year. The present two-year longitudinal study used the IPAW cohort, but was not intended to study intervention effects, which will be reported later. Data were derived from baseline questionnaires and employers' registers of absence for 1919 participants (1305 women, 614 men, mainly in low-skilled jobs) in different occupations from 52 workplaces. These workplaces included municipal care, municipal technical services and a large pharmaceutical company. Analyses were performed by Poisson regression accounting for over-dispersion. After controlling for age, family type, health behaviours and physical work environment variables, we found that high levels of decision authority predicted low absence rates in both women and men. We tested two new psychosocial constructs developed for this study: Predictability (relating to being informed on future events at work) and Meaning of Work (relating to how meaningful and useful the work is perceived to be). Higher Predictability was a significant predictor of lower absence rates in men. This study adds to the body of evidence that the psychosocial working environment influences absence and should therefore be considered to be an important target for intervention.