Network pharmacology and molecular docking methods are used to systematically study the components, targets, pathways and cancer treatment of perilla seeds. With the help of TCMSP, Genecards, and Uniprot databases, the components of perilla seeds, targets, and cancer and tumor targets are collected. Cytoscape software is used to construct a network diagram for the components of perilla seeds, targets, and cancer tumors. The protein is performed through the STRING database interaction network analysis, GO function and KEGG enrichment with the help of DAVID platform. Auto Dock is used to molecularly dock the active ingredients of perilla seeds with anti-cancer and anti-tumor targets. There are 72 targets in total for the anti-cancer and anti-tumor effects of perilla seeds screened, corresponding to 17 effective ingredients. The cancer types involved include bladder cancer, thyroid cancer, pancreatic cancer, etc., among which luteolin and TP53 have the best docking activity. This study builds a network of components-target-disease of perilla seeds, explores the key targets of perilla seeds' anti-cancer and anti-tumor effects, simulates the molecular docking of target proteins and active ingredients, and provides guidance to the clinical application and application of perilla seeds.