Rats have the ability to learn about potential food sources by sampling their odors on the breath of conspecifics. Although this ethologically based social behavior has been transposed to the laboratory to probe nonspatial associative olfactory memory, only a few studies have taken full advantage of its unique features to examine the organization of recently and remotely acquired information. We provide a set of standardized procedures and technical refinements that are particularly useful in achieving this goal while minimizing confounding factors. These procedures, built upon a three-stage protocol (odor exposure, social interaction and preference test), are designed to optimize performance across variable retention delays, thus enabling the reliable assessment of recent and remote memory, and underlying processes, including encoding, consolidation, retrieval and forgetting. The different variants of the social transmission of food preference paradigm, which take a few days to several weeks to perform, make it an attractive and versatile tool that can be coupled to many applications in CNS research. The paradigm can be easily implemented in a typical rodent facility by personnel with standard animal behavioral expertise.