This study identifies conserved protein biomarkers for Autism Spectrum Disorder (ASD) using a cross-species proteomic approach. Proteomic analyses were conducted on serum and brain samples from CNTNAP2 knockout (KO) mice and human ASD patient serum. A total of 132 conserved proteins were identified, with functional enrichment in lipid metabolism, synaptic signaling, and immune response. Machine learning refined these to ten key proteins demonstrating high classification performance. This dataset supports the discovery of ASD biomarkers with potential clinical applications.