Several works have addressed the impact of transport accessibility on regional development. Nevertheless, it is not consensual that new levels of accessibility lead to the improvement of regional development. This article proposes a methodological approach to assess whether significant relationship exist between these two variables. The methodology uses Data Envelopment Analysis to evaluate regional performance and spatial statistics methods (local bivariate relationships) to evaluate the type of relationships between accessibility and performance for each region, along with its significance. The approach was applied to 186 NUTS that cover 19 European countries, extending previous works which focus on global results, in the sense that only one (usually large) area is considered. Results indicate that approximately half the regions have a significant relationship between transport accessibility and regional performance. Logit regression analyses show that central regions are more likely to exhibit significance, as well as regions with high (low) levels of performance surrounded by regions with high (low) levels of accessibility, hinting at a synergy between the two variables. An important policy implication is that regions of high performance surrounded by regions of low accessibility may benefit from investments in the transport infrastructure.