This thesis analyses global agricultural commodity price dynamics, with an emphasis on the causal factors behind movements and their forecastability. The analysis builds upon recent developments in the areas of time series econometrics, agricultural economics and applied economics to provide an empirical examination of agricultural commodity price movements. The main research questions addressed are the following. First, recursive unit root tests are employed to examine whether global commodity prices experienced explosive sub-periods which cannot be explained by underlying economic fundamental movements. Second, a Bayesian Structural VAR is used to model global wheat prices and decompose the causal factors behind price movements. Finally, an examination of the forecastability of agricultural commodity price series is conducted using recently developed dynamic models.