Two demographic/personal characteristics in the Steers and Rhodes' (1978, 1984) process model of employee attendance, and on which considerable research has accumulated in the absenteeism literature, are age and tenure. Recent narrative reviews of this research show that the results are quite mixed and likely to depend upon sex of the sample and absence type. A meta-analysis (Hunter, Schmidt, & Jackson, 1982) of this literature revealed that age (but not tenure) had a modest (p = -.23) relationship with avoidable absences, neither age nor tenure was associated with unavoidable absences, and sex of sample was identified as a moderator. Implications of these findings for the Steers and Rhodes' model are discussed in light of possible directions for future research.