This paper investigates the relationship between Swedish unemployment and labour-force participation. Cointegration analysis supports a robust long-run relationship between the two variables. This finding puts the empirical relevance of the unemployment invariance hypothesis into question.