First paper studies the effects of numbers of trading partners and geographical disadvantage on the agglomeration process in manufacturing sector and the real income distribution among regions. The two configurations of the three-region general equilibrium model with monopolistic competition were developed based on Krugman and Venables’ two-symmetic-region model of 1995. The equilateral triangular configuration model featuring the equal transport costs for all trades enabled the study of the effect of an additional region to the Krugman and Venables’ model. The linear configuration model allowing transport costs between pair-wise trades be not identical enabled the study of the effect of transport cost differences due to geographic location. The symmetric transport costs case suggested that more trading partners leaded to sooner agglomeration. The asymmetric transport costs case, within certain ranges of parametric values, suggested a possibility that agglomeration will take place at the location disadvantage region and become a wealthier region. Second paper studies the occurrence and possible causes of missing trade if found in an experimental data. An individual-choice experiment featuring a partial equilibrium international trade environment was developed around three core international trade theories: Ricardian’s relative technologies, Heckscher-Ohlin’s relative endowment and New Trade Theory models. Experimental data were analyzed based on theoretical profit maximization models and two forms of bounded rational behavior models: the learning and the heuristics. Results indicated that missing trade was also observed in experimental data and for which bounded-rationality in the form of learning and heuristic decision-making seemed to account. The third simulation paper examines the performance of different types of heuristic agents competing against a profit maximizing agent in a partial equilibrium setting with imperfect information. Specifically, the type of competitor is unknown or uncertain to the profit maximizing agent during the process of solving for the optimal strategy. Within the same set of parametric values used in experiment paper, the results showed that profit maximizing agent, whose decisions are undermined by imperfect information, could be outperformed by bounded rational or heuristic agents.