A firm's productivity is composed of two parts: pure technical change and location-specific (agglomeration) externalities. Regional productivity is thus an aggregation of productivity of firms producing similar goods and located in a given region. International trade can affect both components of regional productivity. First, trade openness in a closed economy may alter its internal economic geography. Some regions which become more attractive to firms than before gain an advantage over others from integration into global markets. Second, as a competition pressure, trade liberalization forces the least productive firms to exit, resulting in the growth of aggregate productivity in the industry. The three essays presented in this dissertation explore the relationship between international trade and regional productivity in the presence of heterogeneous firms. In the first essay, a theoretical framework is introduced in order to describe how the above two channels, through which trade affects regional productivity, shape a country's spatial distribution of productivity. Results show that industries, each having its own cost-minimizing location, can be spatially relocated within a country via heterogeneous trade liberalization across industries. Moreover, trade intensifies localization for each industry since most firms in an industry move to or gather around their industry-specific cost- minimizing location. The consequent clustering of firms generates additional localization economies. More importantly, the intensification of localization economies can slow or delay the selection process, i.e. exit of low productivity firms, following trade liberalization. These findings suggest that trade openness induces significant industrial and spatial dynamics (entry, exit and survival) within an economy. The second and third essays are empirical tests on the second channel through which trade openness affects regional productivity using county-level data from Korea and firm-level data from India, respectively. In addition to trade liberalization, regional infrastructure is considered to be another competition pressure for domestic firms, i.e. improved infrastructure in a region induces a similar selection process among firms. These empirical essays investigate the effect of falling trade costs and improving domestic infrastructure on the regional variation of raw productivity using a common methodology. That is, a spatial econometric procedure is applied to a production function framework to estimate total factor productivity (TFP) by region and industry, while controlling for potential external and spatial effects. The mean and alternative percentiles of the regional raw productivity distribution are then specified as functions of international and domestic competition indicators. International competition is represented by trade costs, which are estimated as frictions in a gravity-type trade model, while road density is considered to capture the level of a region's infrastructure. In both Korea and India, it is found that trade costs reduction significantly shifted to the right, particularly the 10th percentile value of, the regional productivity distribution. However, a change in the level of infrastructure appears to bring about a higher change in regional productivity relative to a change in the international competition level. Therefore, the relative contribution of trade costs and infrastructure to regional productivity should be evaluated with attention to the costs underlying these options for regional development.