In this paper, we provide causal evidence that firms serve new markets which are geographically close to their prior export destinations with a higher probability than standard gravity models predict. We quantify the impact of this spatial pattern using a data set of Chinese firms which had never exported to the EU, the United States, and Canada before 2005. These countries imposed import quotas on textile and apparel products until 2005 and experienced a subsequent increase in imports of previously constrained Chinese firms. Controlling for firm-destination specific effects and accounting for potential true state dependence we show that the probability to export to a country increases by about two percentage points for each prior export destination which shares a common border with this country. We find little evidence for other forms of proximity to previous export destinations like common colonizer, language or income group.