Using hierarchical Bayesian estimation of RT distributions, we present a novel application of Systems Factorial Technology (Townsend & Nozawa, 1995) to the retrieval of item and associative information from episodic memory. We find that item and associative information are retrieved concurrently, with positive memory evidence arising from a holistic match between the test pair and the contents of memory, in which both item and associative matches are pooled together into a single source. This retrieval architecture is inconsistent with both strictly serial processing and independence of item and associative information. Pooling of item and associative matches implies that while item and associative information may be separable, they are not qualitatively different, nor are qualitatively different processes (e.g., familiarity vs. recollection) used to retrieve these kinds of information.