The calibration of the tertiary mirror of the Thirty Meter Telescope, also known as the giant science steering mirror (GSSM), is a step of great significance during its testing process. Systematic, drift, and random errors constitute the major limitations to the accuracy of the calibration measurements. In this article, we estimated the errors in the calibration of the GSSM with a laser tracker. For the systematic error, a measurement strategy based on the standard bar method was successfully designed and applied. At the same time, we can distinguish between the drift and random errors by means of a correlation analysis. The systematic error, which depends strongly on the configuration of the system formed by the GSSM and the laser tracker, was estimated to be 20 μm for the GSSM prototype. The random error, averaging 15 min, was about 4 μm. The correlation coefficients among three different noise measurements are all lower than 0.1, which indicates that the noise is dominated by random errors. Finally, the error can be sufficiently suppressed by rearranging the position of the spherically mounted retroreflectors. The result shows that the accuracy of the measurement can be improved by 21.4% with the new arrangement method.