Utilizing a comprehensive dataset for Turkish manufacturing firms, we analyse differentials in the post-entry effects on productivity of exporting to markets with different income levels. We employ propensity score matching techniques with multiple treatments, together with a differences-in-differences (DiD) methodology. Controlling for firm level mark-ups, we explore whether the post-entry effects on productivity are driven by changes in physical productivity. The results confirm the learning-by-exporting hypothesis, and suggest physical productivity gains, in particular for exports to high income countries as opposed to middle low-income countries, even after controlling for the composition of exports. This suggests that where a firm export does matter for productivity growth.