Preferential trade agreements (PTAs) have proliferated over the past 60 years. While a small number of recent studies have examined empirically the economic determinants of the likelihood of a pair of countries having a PTA, this study explains empirically the timing of all PTA formations and enlargements from 1950 through 2006 using duration analysis. Our main and novel goal is to predict (in- and out-of-sample) a substantive share of these 1,560 PTA events using a parsimonious model with mainly economic variables, taking selection dynamics into account. Our analysis reveals that we can predict correctly in-sample the actual year of entry into force for 26% of the 1,560 bilateral PTA formations/enlargements in the period 1950-2006 among 10,518 pairings of 146 countries using only a few economic and political variables. Moreover, we can predict correctly in-sample 57% of these PTA events within a 10-year window leading up to the event using this model. The model also performs well out-of-sample for the near term (82%), but not if the out-of-sample period is very long. We conclude with an evaluation of the model's ability to predict the timing of the North American Free Trade Agreement, the European Union's formation and enlargements, and the model's ten most likely post-2006 PTA events. (JEL F14, F15), I. INTRODUCTION One of the most notable economic events since World War II is the proliferation of preferential trade agreements (PTAs), including free trade agreements (FTAs) and customs unions (CUs). [...]