Between the years 1996 and 2000, over 2000 projects were carried out in Finland with the aim of finding innovative measures for crossing the job threshold. Among them was the Pathway-to-Work Project, which aimed at tailoring return-to-work plans for 140 middle-aged, long-term unemployed participants with various disabilities and getting half of them into work or training. This study of the Pathway-to-Work Project had two research objectives. First, to evaluate the outcomes of the return-to-work rehabilitation project and second, to determine what combination of different measures seemed necessary and effective in the rehabilitation of long-term unemployed people with disabilities. The research design comprised three parts: a quantitative quasi-experimental part with a matched control group, a register follow-up and the collection of qualitative data. The main variables used to evaluate the outcomes were (1) the changes in the labour market situation during the 2-year register follow-up, (2) the changes in distress (measured by the General Health Questionnaire-12), perceived competence (measured by Wallston's Self-Performance Survey) and sense of coherence (measured by Antonovsky's SOC-13) during the intervention and (3) the description of the process in the project. In the 1-year follow-up, 31% of the participants were found to be at work and 37% unemployed. In the 2-year follow-up, 14% were at work and 59% unemployed. The jobs seemed to be subsidized for a period of half a year to a year. The difference between the project group and the matched control group was remarkable: at the end of the project, only 9% of the control group were at work and 86% unemployed. The participants' distress level decreased remarkably and their perceived competence increased, but their sense of coherence did not change. The results showed that even carefully tailored client work enables only some of the long-term unemployed people with disabilities to cross the job threshold and that other means of policy, strategy and intervention are needed to link the return-to-work interventions more closely with work, work places and enterprises.