Gene mutations are important biomarkers for the diagnosis, classification, monitoring, and prognosis evaluation of cancers and genetic diseases. Both personalized cancer treatment and noninvasive prenatal testing require methods to accurately determine the abundance of mutation. At present, the widely adopted and convenient methods for measuring mutation abundance are mainly based on relative quantification, which requires negative samples and strict control of the analyte amounts. The development of DNA-probe-based methods that can determine the mutation abundance without negative samples nor control of analyte amount is highly preferred. The key to solving this bottleneck lies in whether the probe's response to mutation abundance can be completely independent of the number of targeted DNA strands. Herein, we propose the design of a self-internal-reference probe system. We established a theoretical model of this system and used the model to guide the design of probes. In this model, we provided quantitative corrections to the test results from the internal reference, thereby eliminating the influence of substrate amount. Therefore, the purification and quantification processes toward polymerase chain reaction (PCR) amplicons can be omitted. We applied this system to analyze unquantified PCR products aimed at cancer mutation detection and noninvasive prenatal testing.