SHANK3 gene is a highly replicated causative gene for autism spectrum disorder (ASD) and has been well characterized in multiple Shank3 mutant rodent models. When compared to rodents, domestic dogs are excellent animal models in which to study social cognition as they closely interact with humans and exhibit similar social behaviors. Using CRISPR/Cas9 editing, we recently generated a dog model carrying Shank3 mutations, which displayed a spectrum of autism-like behaviors, such as social impairment and heightened anxiety. However, the neural mechanism underlying these abnormal behaviors remains to be identified. We used Shank3 mutant dog models to examine possible relationships between Shank3 mutations and neuronal dysfunction. We established a protocol for the electrophysiological recording of canine brain slices and revealed that excitatory synaptic transmission onto PFC layer 2/3 pyramidal neurons in Shank3 heterozygote dogs was impaired, and this was accompanied by reduced dendrite complexity and spine density when compared to wild-type (WT) dogs. Postsynaptic density (PSD) structures were also impaired in Shank3 mutants, however, pyramidal neurons exhibited hyperexcitability. Our study demonstrated the feasibility of using canine brain slices as a model system to study neuronal circuitry and disease. Shank3 haploinsufficiency causes morphological and functional abnormalities in PFC pyramidal neurons, supporting the notion that Shank3 mutant dogs are new and valid animal models for autism research.