Transitions from unemployment into temporary work are often succeeded by a transition from temporary into regular work. This paper investigates whether temporary work increases the transition rate to regular work. We use longitudinal survey data of individuals to estimate a multi-state duration model, applying the ‘timing of events’ approach. The data contain multiple spells in labour market states at the individual level. We analyse results using novel graphical representations, which unambiguously show that temporary jobs shorten the unemployment duration, although they do not increase the fraction of unemployed workers having regular work within a few years after entry into unemployment.