In recent decades, one of the most notable developments in social policy has been the expansion of active labour market policies (ALMPs): training schemes that aim to speed up unemployed people’s return to the labour market. At the same time, academic and political attention has begun to rethink the traditional ways in which social policies like ALMPs should be evaluated: away from typically economic-oriented outcomes towards health and social indicators, such as subjective well-being and social capital. This has led to an emerging argument that ALMPs can be used to improve the health and social environment of unemployment, which decades of research has shown to be associated with a wide range of deleterious outcomes. This paper tests this argument by analysing longitudinal data from the long-running British Household Panel Survey and its successor Understanding Society. The results show that relative to open unemployment, ALMP participation is associated with increased well-being amongst the unemployed, although there is no effect on health or social capital.