The need for better targeting of active labour market programmes is evident from the many evaluation studies that find insignificant or even negative effects. A statistical system could contribute to a more precise targeting of labour market programmes to those individuals who are likely to benefit from them. Such a system could assist caseworkers in selecting adequate programmes on an individual basis. In this paper, international experiences with these systems are surveyed and potential approach for Switzerland is developed. The simulated outcomes indicate that a statistical selection system could have contributed to a substantial re-employment increase.