The inherently multivariate nature of functional brain imaging data affords the unique opportunity to explore how anatomically disparate brain areas interact during cognitive tasks. We introduce a new method for characterizing inter-regional interactions using event-related functional magnetic resonance imaging (fMRI) data. This method's principle advantage over existing analytical techniques is its ability to model the functional connectivity between brain regions during distinct stages of a cognitive task. The method is implemented by using separate covariates to model the activity evoked during each stage of each individual trial in the context of the general linear model (GLM). The resulting parameter estimates (beta values) are sorted according to the stage from which they were derived to form a set of stage-specific beta series. Regions whose beta series are correlated during a given stage are inferred to be functionally interacting during that stage. To validate the assumption that correlated fluctuations in trial-to-trial beta values imply functional connectivity, we applied the method to an event-related fMRI data set in which subjects performed two sequence-tapping tasks. In concordance with previous electrophysiological and fMRI coherence studies, we found that the task requiring greater bimanual coordination induced stronger correlations between motor regions of the two hemispheres. The method was then applied to an event-related fMRI data set in which subjects performed a delayed recognition task. Distinct functional connectivity maps were generated during the component stages of this task, illustrating how important and novel observations of neural networks within the isolated stages of a cognitive task can be obtained.