This paper examines the long-term earnings losses of displaced workers in Portugal, using a nationally representative longitudinal linked employer-employee data set. The results show that four years after displacement the earnings of displaced workers remain around 9% (women) to 12% (men) below their counterfactual expected levels. The post-displacement earnings losses are mainly associated with the loss of tenure within the firm and, to a lesser extent, to the loss of sector-specific features. Furthermore, workers who experienced a spell of nonemployment are the most affected by job displacement. Finally, this study points to the importance of controlling for employers' characteristics in this type of wages-dynamic analysis, since there are systematic differences in earnings between displaced and non-displaced workers that stem from differences in firm characteristics. Ignoring them may confound the evaluation of the earnings losses.