Background: With the rapid progress of genetic engineering and gene therapy methods, the World Anti-Doping Agency has raised concerns regarding gene doping, which is prohibited in sports. However, there is no standard method available for detecting transgenes delivered by injection of naked plasmids. Here, we developed a detection method for detecting transgenes delivered by injection of naked plasmids in a mouse model that mimics gene doping. Methods: Whole blood from the tail tip and one piece of stool were used as pre-samples of injection. Next, a plasmid vector containing the human erythropoietin (hEPO) gene was injected into mice through intravenous (IV), intraperitoneal (IP), or local muscular (IM) injection. At 1, 2, 3, 6, 12, 24, and 48 h after injection, approximately 50 µL whole blood was collected from the tail tip. One piece of stool was collected at 6, 12, 24, and 48 h. From each sample, total DNA was extracted and transgene fragments were analyzed by Taqman quantitative PCR (qPCR) and SYBR green qPCR. Results: In whole blood DNA samples evaluated by Taqman qPCR, the transgene fragments were detected at all time points in the IP sample and at 1, 2, 3, 6, and 12 h in the IV and IM samples. In the stool-DNA samples, the transgene fragments were detected at 6, 12, 24, and 48 h in the IV and IM samples by Taqman qPCR. In the analysis by SYBR green qPCR, the transgene fragments were detected at some time point in both specimens; however, many non-specific amplicons were detected. Conclusions: These results indicate that transgene fragments evaluated after each injection method of naked plasmids were detected in whole-blood and stool DNA samples. These findings may facilitate the development of methods for detecting gene doping.