Aiming at the problem that the high shear rate during micro-injection molding leads to a rise in melt temperature due to viscous dissipation, which further affects the accuracy of viscosity measurement, in this paper an average-value method based on measuring temperature rise from viscous dissipation to correct the measured viscosity was proposed, so as to obtain the real rheological properties of polymer melts at the microscale. A twin-bore capillary rheometer was used to measure the viscosity and viscous dissipation temperature rise of polypropylene (PP) at different shear rates in a capillary die with a diameter of 500 μm. The viscosity was corrected, and the numerical simulation was compared with the experimental pressure for verification. The results show that the simulated pressure based on the corrected melt viscosity is closer to the experimental pressure. At the three inlet temperatures of 220 °C, 240 °C and 260 °C, the average relative error between the corrected simulated pressure and the experimental pressure is 2.83%, with a maximum local error of 4.82%; the average relative error of the uncorrected simulated pressure is 9.24%, with a maximum local error of 15.81%.