Objectives: Research on the correlation of unemployment and health has a long tradition. This correlation is explained by two opposite hypotheses: The causality theory proposes the occurance of health impairments as a consequence of unemployment; according to the selectivity theory, however, psychologically impaired persons are more likely to become unemployed than healthier individuals. Methods: Both hypotheses are examined by analysing the three waves of the Saxonian profile study (1991, 1996, 2004). The sample contains N = 153 persons that were interviewed at all three waves. The average age increased from M = 18 years in 1991 to M = 31 in 2004. 57% of the sample were female. The subjective psychological distress was measured by a newly designed 4-item instrument, the D-Score, which was proven to be reliable and valid. In addition, experiences with unemployment were acquired. Results: In 1996 nearly 50% of the respondents had already experiences with unemployment, in 2004 nearly 60%. The duration of unemployment was associated with the D-Score within the individual waves: Persons, who were unemployed for a longer time, were more distressed in each case. The psychological condition from earlier waves was strongly correlated with the total period of unemployment experienced, which was raised in 2004. Conclusion: The results give hints for both causality and selection hypotheses. More distressed persons may become unemployed more frequently, and unemployment may in turn lead to more pronounced distress. Thus, an interplaying, self-energizing effect of selection and causality may be assumed in the sense of a vicious circle: Psychologically impaired persons are more often unemployed and then suffer more from the consequences of their unemployment. This in turn may lead to poor chances on the job market.