This paper investigates short and long-run effects of trade liberalization on employment and wages. Employment and wage equations are estimated using data (1971?96) for importable and exportable sectors in Tunisia. Causality tests show that causality is unidirectional. Wages strongly causes employment but employment does not cause wages. There is significant difference in the direction of responses in the short and long run. Results from empirical testing using the models find only support for the short-run theoretical predictions for the exportable sector. Similar results obtained for the importable sectors. We find the differences in the short and long-run wage and employment responses to changes in export to be explained by learning by doing, organizational changes and improved factor utilization and labour productivity. A possible reason for the divergence of theory and practice is that the theoretical model is premised on the basis of a fixed supply of labour. Exportable employment could therefore only rise if importable employment fell. However, as we have seen, the supply of labour increased dramatically in Tunisia as women entered the labour market. This allowed importable employment to be maintained (even slightly increased) as the exportable sector expanded.