Autophagy, a cellular process where cells degrade and recycle their own components, has garnered attention for its potential role in psychiatric disorders, including schizophrenia (SCZ). This study aimed to construct and validate a new autophagy-related gene (ARG) risk model for SCZ. First, we analyzed differential expressions in the GSE38484 training set, identifying 4,754 differentially expressed genes (DEGs) between SCZ and control groups. Using the Human Autophagy Database (HADb) database, we cataloged 232 ARGs and pinpointed 80 autophagy-related DEGs (AR-DEGs) after intersecting them with DEGs. Subsequent analyses, including metascape gene annotation, pathway and process enrichment, and protein-protein interaction enrichment, were performed on the 80 AR-DEGs to delve deeper into their biological roles and associated molecular pathways. From this, we identified 34 candidate risk AR-DEGs (RAR-DEGs) and honed this list to final RAR-DEGs via a constructed and optimized logistic regression model. These genes include VAMP7, PTEN, WIPI2, PARP1, DNAJB9, SH3GLB1, ATF4, EIF4G1, EGFR, CDKN1A, CFLAR, FAS, BCL2L1 and BNIP3. Using these findings, we crafted a nomogram to predict SCZ risk for individual samples. In summary, our study offers deeper insights into SCZ’s molecular pathogenesis and paves the way for innovative approaches in risk prediction, gene-targeted diagnosis, and community-based SCZ treatments.