LISREL estimates of the parameters of a causal model of absenteeism are reported for a sample of 425 full‐time employees of a 327‐bed medical centre. The model modifies and extends the conceptual framework of Steers&Rhodes (1978) and includes routinization, centralization, pay, distributive justice, work involvement, role ambiguity, conflict and overload, kinship responsibility, organizational permissiveness, job satisfaction, job involvement, organizational commitment, health status and alcohol involvement as the determinants of absenteeism. Based on self‐report measures of absence frequency, significant positive direct effects of kinship responsibility, organizational permissiveness, role ambiguity and alcohol involvement, and negative direct effects of centralization, pay and job satisfaction yielded anR2for absenteeism of 21.6. Job satisfaction completely mediated the effects of routinization and work involvement, and partially mediated the effects of centralization and role ambi