In this thesis, we consider the trading world economy with some firms exhibiting increasing returns to scale technology. We focus on the problems of existence of equilibrium, gains from trade and the effects of increasing returns in developing countries. The dissertation consists of three independent essays. The first essay proves the existence of a world trade equilibrium under a special pricing rule--marginal cost pricing. Our results support an internationally decentralized subsidy arrangement. In addition, we show that the equilibrium existence results of Debreu (1959) and Grandmont and McFadden (1972) can be derived as a special case of ours. The second essay shows that for a small country there exists at least one free trade marginal cost equilibrium which is Pareto non-inferior to autarky regardless of variable returns to scale. The result can be applied to three kinds of economies: a centrally planned economy with one consumer; a multi-consumer economy with lump-sum transfers; and a multi-consumer economy without lump-sum transfers but with feasible commodity taxes. The paper reports a generalization of the result of Dixit and Norman (1986) on the gains from trade without lump-sum compensation. The third essay deals with the effects of increasing returns on developing countries. We find that a rapidly growing developing country which is oriented toward the rest of the world may exhibit strongly increasing returns. Moreover, it may enjoy comparative advantage in the increasing returns sectors. The policy implication is that developing countries should extend the industries with increasing returns in order to enjoy the gains from economies of scale.