Dscam2 is a conserved transmembrane protein that plays a critical role in the development of the Drosophila visual system. Dscam2-Dscam2 homophilic interactions induce repulsion between neighbouring neurons, compartmentalising the brain by creating boundaries between different regions. Dscam2 also has distinct post-developmental functions where it regulates synaptic strength through a PI3K pathway that decreases the number of neurotransmitter-containing vesicles at synapses. Here, we investigate the intracellular trafficking compartments involved and whether these post-developmental functions regulate higher order processes such as memory. To address whether vesicle trafficking is perturbed in the Dscam2 mutants, we used electron microscopy in combination with GFP-tagged Rab proteins and APEX-GBP to compare Rab vesicle distributions in larval motor neurons. We found that large vesicles containing Rab11 were abundant in the Dscam2 mutant but rare in controls. Light microscopy confirmed an increase in Rab11 protein in Dscam2 mutants and mRNA levels were not different from controls. To determine whether this Rab11 phenotype was dependent on PI3K signalling, we performed rescue experiments. Both vesicular and plasma membrane classes of PI3K were able to significantly reduce Rab11 levels in the Dscam2 mutants. Thus, Dscam2 promotes PI3K signalling that regulates Rab11 vesicle distribution and leads to fewer synaptic vesicles at synapses. Given that synaptic plasticity is critical for memory formation, we asked whether Dscam2 played a role in this process. We used a courtship conditioning assay whereby a male paired with a pre-mated, non-receptive female learns that his mating attempts are futile in a training session and then recalls this information in a subsequent testing session. A memory index is then calculated based on the reduction in courting in the testing session. Homozygous Dscam2nullmutants and animals that lacked Dscam2 in the learning and memory centre of the brain, both exhibited a defect in memory formation using this assay. To determine whether Dscam2 was required for the development of the memory circuitry or for the regulation of the adult circuitry, we conditionally knocked it down in adults and found a similar memory defect, suggesting that this was a post-developmental function for Dscam2. To link our trafficking data in motor neurons with the memory phenotype, we tested whether PI3K could rescue the memory defect. Similar to the Rab11 rescue experiments, both vesicular and plasma membrane classes of PI3K rescued the Dscam2 mutant memory defect. These data indicate that Dscam2 regulates trafficking pathways in adults required for learning and memory.