The aim was to investigate whether a preventive intervention carried out in a predominantly female workplace, that of hospital cleaners (consisting of a group of 97 women), had any effect on patterns of absenteeism. As a background, a model for analysing complex patterns of absenteeism, including sickness absences, was also developed. A further aim was to study the interactions between different forms of absenteeism. Comparison was made with a reference group consisting of employees in the same job category who only received the customary personnel support. For individuals in the intervention group who were <42 years of age, total absence due to sickness decreased. In a multiple regression analysis, the contribution from the intervention to the decrease was significant at the 5% level. This change was particularly obvious in those who had a previous history of high absence due to sickness. No clear relationship was shown between short-term absenteeism and the interventions applied. For those who were > 42 years, short-term absence decreased for those who had been in the same jobs for a long time. The combination of increased age and experience showed a tendency to enhance this decline in short-term absenteeism due to sickness. For those > 42 years, and who at the same time have a previous history of high absenteeism, long-term absenteeism due to sickness seemed to be increasing. Increased experience tended to reduce this increase in long-term sickness absence. This combination of different effects possibly indicated the presence of a process of selection which determined who remained in the job as opposed to those who did not. An important conclusion is that different forms of absenteeism need to be looked at in parallel, and at the same time multivariate statistical analysis needs to be carried out to determine the different interactions between the factors.