<div>The number of preferential trade agreements (PTAs) has increased tremendously since 1990. The natural question to ask is why. PTAs are not only about lowering down tariffs further than the most favoured nation (MFN) tariff levels. There are many economic and non-economic policies other than border policies that are addressed in PTAs. Trade agreements dealing with border policies (tariffs) are referred to as “shallow”; and those that are dealing with a broader set of policies are referred to as “deep”. Therefore, PTAs are about something deeper. Parallel to the increase in the PTAs, trade in intermediate inputs has grown exponentially over past decades. Therefore, the first question that arises is whether trade in intermediate inputs generates the need for deep integration. In this dissertation, we show that the nature of trade in intermediate goods requires deep integration. The second question to be addressed is whether the deep trade agreements need to be preferential. With a three-country model, we show that the deep bilateral agreements are rarely chosen over the shallow agreements. Finally, by introducing the deep integration in the multilateral trading system, we conclude that although trade in intermediate inputs calls for deep integration, they do not call for deep PTAs. However, deep integration is better implied under multilateral agreements. Therefore, the deep integration in economic policies does not contribute to the increase in the number of PTAs.</div>