In the present study, a diagnostic method using a first-principles based digital twin is proposed for application during engine verification test. The first-principles model is applied to generate the digital twin model using a small number of measurement data. A method is developed to identify abnormalities of components in a short time and to classify faults through detailed inspection. The digital twin model is generated through performance adaptation with a basic first-principles model. Performance adaptation factors are used for compressor, secondary air system, nozzle, exhaust gas temperature, and engine thrust. Gas path analysis is conducted using the digital twin and the performance test data in real time. Health parameters for compressor capacity and efficiency, secondary air system, and nozzle are calculated from the gas path analysis. Multi-variable Newton-Raphson method is employed for the performance adaptation and the gas path analysis in real time. A single-shaft gas turbine engine is utilized to test and evaluate the effectiveness of the proposed method. Three tests are performed for engine verification. The diagnostic method is applied to the engine test. As a result, it is confirmed that the proposed method can simultaneously detect abnormalities in HPC VGV, secondary air system, and variable area nozzle.