When information about workers'' abilities is imperfect, past employment histories are important determinants of unemployment duration. This paper studies the effects of employment protection laws on unemployment spells in two countries: Italy and Spain. Italy and Spain''s adoption of fixed-term contracts with lower firing costs in the mid 1980s provides an interesting empirical framework to study the effects of policies affecting firing costs. The evidence presented supports the hypothesis that when firing costs increase, so does the stigma attached to bad employment histories. First, I show that in Italy people, who became unemployed because of termination of the employment contract, have 10 per cent lower probability per year to remain unemployed than fired workers. Second in Spain, workers dismissed from a temporary job have 30 per cent shorter unemployment spells than workers fired from regulator contracts.