The overall aim of this thesis was to identify factors that early in a sick leave period could illuminate people’s resources and obstacles for returning to work; factors that could predict need for professional support in the sick leave process leading to a positive outcome for the individual. This thesis consists of four studies. The studied population was living in a rural part of a northern county in Sweden, and had been on sick leave for 28 days irrespective of diagnosis. The participants have been followed for five years with different investigation methods and subgroups of participants. The thesis was initiated by sending a questionnaire to the participants; the responses were compared to groups of healthy controls. The aim was to find predictors of the likelihood of a return to work that can be identified early in a sick leave period. The main focus was on individual mental resources (Study I). The next phase involved studies of sick leave statistics obtained from the Swedish Social Insurance Office, for the study population two years before and two years after the start of the research project. The material from the sick leave statistics was analysed together with responses from the questionnaire study with Partial Least Square (PLS). This was done to search for connections between the participants and factors of concern for sick leave and returning to work in different subgroups (Study II). In-depth interviews also were conducted to explore what the participants experienced as important in their sick leave process as they progressed back to work, or to long term sick leave or disability pension (Study III). Another study investigated client files in the Swedish Social Insurance Office to describe what information that was possible to find: in terms of medical and vocational rehabilitation including assessments, predictors for the outcome of the sick leave process, and the quality of the information in the files (Study IV). The findings showed significantly lower life satisfaction and psychosocial resources in the study-group compared to the healthy controls. The PLS analysis showed that it was the impact of multiple factors that influenced the study group, and the PLS analysis could help with early prediction of the outcome. Important factors were: personal belief in an ability to work in the future, number of sick leave days in the past, diagnosis, self-evaluated symptoms, life satisfaction and sense of coherence, length of education and sector of employment, and many different consequences in daily life caused by activity limitations. The interview study confirmed these results and added the following important resources: confirmation and support, structural and contextual factors, and participation in the sick leave process. Ideal-types were crystallized that can be identified in the early phases of the sick leave process. The client file study showed that some information was possible to find but a majority of the wanted information was limited why an improvement on the quality of documentation is suggested to give better basis for the files.