Drug‐metabolizing enzymes play an important role in the metabolism of drugs in vivo. Their activity is an important factor affecting the rate of drug metabolism, which directly determines the intensity and persistence of drug action. Patients taking medication can be divided into different metabolic types through detection of CYP2C19 drug‐metabolizing enzyme gene polymorphisms, which can then be used for medication guidance for clopidogrel. Here, we describe a detection method based on real‐time polymerase chain reaction. This method uses multicolor melting curve analysis to accurately identify different mutation sites and genotypes of CYP2C19 * 2, CYP2C19 * 3, and CYP2C19 * 17.The detection limit of plasmid samples was 1 copies/μl; that of genomic samples was 0.1 ng/μl. The system can detect nine types of CYP2C19 * 2/3/17 at three sites in one tube, quickly achieving detection within 1 h. Combined with the sample release agent, sample extraction was completed in 5 s, achieving rapid diagnosis without extraction for timely diagnosis and treatment. Furthermore, the system is not limited to blood samples and can also be applied to oropharyngeal and saliva samples, increasing sampling diversity and convenience. When using clinical blood samples (n = 93), the detection system we established was able to quickly and accurately identify different genotypes, and the accuracy and effectiveness of the detection were confirmed by Sanger sequencing.Due to its accuracy, rapidity, simple operation and low cost, detection technology based on real‐time polymerase amplification combined with melting curve analysis is expected to become a powerful tool for detecting and guiding clopidogrel use in countries with limited resources.This article is protected by copyright. All rights reserved