The perception and recognition of odors allow animals to identify food sources, avoid predators, and find potential mates as they navigate their surroundings. These behaviors require at least three essential compu-tations: (1) the ability to discriminate between similar odors while simultaneously allowing generalization for odor recognition; (2) acquiring and retrieving odor associations; and (3) assigning value to salient odors to guide appropriate behaviors. Although the piriform (PCx) cortex is thought to be the locus for odor perception and odor learning and memory, whether or how the PCx performs these essential computations remains unknown. Our long-term goal is to mechanistically understand how odor information is processed to drive appropriate behaviors. The objective of this proposal is to identify where and how the PCx performs these three essential computations. The PCx has two major types of principal cells, semilunar cells (SLs) and py- ramidal cells (PYRs), which have distinct morphologies, synaptic connectivity, and projection targets. Our central hypothesis is that different PCx cell types play distinct and specific roles in these essential com-putations. To this end, we propose the following three aims. Aim 1: To reveal the neural substrates of discrimination and generalization in PCx. We will train mice to discriminate between two odorants and then record from identified SLs and PYRs as the mice are pre-sented with a morphing series of binary mixtures of the two. Our preliminary data support our hypothesis that SLs discriminate and PYRs generalize responses to similar odors. Aim 2: To reveal the neural substrates for acquiring and retrieving odor memories. We do not under-stand how odor memories are formed, or where and how are they stored and retrieved? We will use viral tools with different transgenic mouse lines to selectively suppress activity in SLs or PYRs just before the acquisition or before the retrieval phases of an odor memory task. We hypothesize that SLs are required for the acquisi- tion but not the retrieval of odor memories, while PYRs are required for both acquisition and retrieval. Aim 3: To determine how odor identity and value information emerge in the medial prefrontal cortex (mPFC). mPFC receives direct projections from PCx and encodes odor value. We hypothesize that odor information is directly routed from PCx to mPFC by one specific subset of PYRs (Tbr1 neurons); that Tbr1 neurons encode odor identify but not odor value; and that the conjunction of odor identity and odorvalue in mPFC occurs through strengthening connections between PCx Tbr1 neurons and their postsynaptic partners in mPFC. This study will reveal the distinct roles that different subsets of PCx neurons and circuits play in processing odor information to guide flexible and appropriate behaviors, and can provide more general insight into how information is transformed within and across canonical circuits throughout the brain.