Terms and theories of work attendance vary according to their use and focus. This paper analyzes long-term work attendance in relation to social, psychosocial, and health-related factors. Register-based and questionnaire-based data covering 3,804 human service organisation workers over a three-year period were analyzed at individual and work-unit level. The results showed positive relationships between work attendance and male gender, high income, work commitment, job satisfaction, and having positive feelings towards work. High work attendance combined with work commitment, stress, or pain did not show any negative long-term effects upon short-term or long-term sick leave. Instead, work attendance seemed to be more associated with stable patterns of behaviour. Register-based measures of work attendance (at most 4–7 days of sick leave per worker per year) may be a useful tool in managing psychosocial work environment and related behaviour, but their inability to encompass information regarding individual health and disease must be borne in mind.