Stationary job search theory is used as a starting point for an empirical analysis of search durations. Data have been derived from the Dutch Socio-economic Panel and cover the period 1985-87. They are particularly interesting since the reservation wage and the conditional expected wage are observed. This makes it possible to analyze the relation between unemployment benefit, reservation wage, and search duration. It appears that other factors, such as gender and educational level, influence duration far more than the benefit and the reservation wage. A close connection is found between benefit and reservation wage (elasticity of 0.3). Copyright 1993 by Blackwell Publishing Ltd