Paleomagnetically derived apparent polar wander paths form the foundation of quantitative plate tectonic reconstructions. However, deformation leading to vertical axis block rotations displaces paleomagnetic poles away from their original positions, leading to an ambiguity as to which pole, or group of poles, best approximates the “true” reference pole position for a given time. Here we show that the best estimate of the “true” reference pole will match the observed paleolatitude (pλ) for each point on the plate. This means that the expected pλ from the “true” reference pole minus the observed pλ, derived from each individual study, will average to zero. Histogram plots and associated parameters help further discriminate between candidate reference poles when more than one of them fulfills the zero-average requirement within prescribed uncertainty limits. Our analysis of 44 Late Permian to Middle Triassic paleomagnetic poles from the South China plate corroborates previous assumptions that the poles from Sichuan Province best represent the “true” reference for the South China plate. To better understand the age of the rotations, we studied the paleomagnetism of Lower and Upper Triassic rocks from the Shiwandashan region in Guangxi Province. Early Triassic paleomagnetic directions isolated at high temperature demagnetization steps are of dual polarity and pass the fold test. This magnetization component is indistinguishable at 95% confidence limits from Middle Triassic paleomagnetic directions from other parts of Guangxi Province. The corresponding pole for this component lies within the swath of Early to Middle Triassic paleomagnetic poles from the South China plate, confirming that Guangxi has been a part of the South China plate since at least the Early Triassic. Late Triassic paleomagnetic data require further study before their complex magnetizations can be interpreted.