In the late Ming Dynasty, European astronomy was first systematically introduced into China through Jesuit missionaries, culminating in the compilation of the Chongzhen lishu (崇禎曆書, Chongzhen Reign Treatises on Calendrical Astronomy), wherein a system of calendrical astronomy was built up on the basis of the works by European astronomers from Ptolemy to Copernicus, Tycho and even Kepler. Since solar motion is very important in a system of calendrical astronomy, it underwent continuous revisions from the Chongzhen lishu to its two later re-editions, Xiyang xinfa lishu (西洋新法曆書, Treatises on Calendrical Astronomy According to New Methods from the West) and Lixiang kaocheng (曆象考成, Thorough Investigation of Calendrical Astronomy). This paper first analyses the overall differences of the three versions of the solar theory, and then concentrates on the solar models used in the whole series and explores their rationalities. Finally, the paper examines the change of the basic parameters in the solar theory and their effects on the accuracy, and analyzes the accuracy of the three versions of solar theory. Our results indicate that although the continous revisions made the solar theory more coherent, they contributed little to the improvement of the actual accuracy of the theory itself. On the other hand, however, they reflect the Jesuit efforts in making the system more perfect to ensure its superiority over the rivaling Chinese and Islamic systems.