Herbicide residues in agricultural products can have adverse effects on the environment and human health, therefore, there is an urgent need to establish a sensitive, rapid, and wide-ranging detection method. In this study, haptens of phenylurea herbicides (PUs) and sulfonylurea herbicides (SUs) were analyzed and designed based on computational simulation techniques, and two high-performance broad-spectrum monoclonal antibodies against PUs and SUs were prepared. On this basis, a multi-colloidal gold immunochromatography assay (multi-CGIA) was developed to simultaneously detect 13 herbicides in wheat. The visual limit of detection (vLOD) for PUs including diuron, chlortoluron, neburon, chlorbromuron, and linuron was 1–2 μg/kg. The vLOD for SUs including metsulfuron methyl, ethametsulfuron-methyl, sulfometuron-methyl, tribenuron methyl, cinosulfuron, triasulfuron, chlorimuron-ethyl, and chlorsulfuron was 2–10 μg/kg. The results of real sample determination indicated that the multi-CGIA is accurate, stable, and reliable, and adaptable to on-site preliminary screening of actual samples.