BasedonthedatafromChina'slistedtextilecompanies duringtheperiodfrom2001tO₂010，we empirically test the two hypotheses of new-newtrade theory. The results indicate that(1)exporters have higher productivity than non-exporters and those with higher export intensityare more productive than those with lower intensity；(2)the self-selection effect does exist，i.e. there is acausal relationship be-tween productivity and export，and that the effect is more significant among exporters with higher export intensity；(3)for the sample companies as awhole，learning-by-exporting does not exist；but，forthe companies with high export intensity，that effect does exist. We conclude that the empirical test based on China's firm level data during the past 10～15years should further take firm-specific export barriersinto account，besides such special factors as processing trade mode and ownership.