Objective To investigate the relationship between compatibilities of prescriptions for treating stroke and their dynasty and resource distribution, provide decision support for clinical applications. Methods Totally 905 prescriptions were selected from Database of Chinese Medicine Prescriptions. Information of selected prescriptions was standardized. The frequency of single herbs appeared in the prescriptions was counted. The core herb groups of the prescriptions were calculated by Apriori algorithm, and categorized by different dynasties. The concept of classical value of prescriptions (CVP) was proposed to characterizing the degree of classical of prescriptions and revealing the law of development of the prescriptions for treating stroke. Results The selected prescriptions involved 1479 different herbs, which were picked from 245 medical books. The top 3 most frequently used herbs were Saposhnikoviae Radix (40.22%), Angelicae Sinensis Radix (36.69%) and Chuanxiong Rhizoma (32.71%). Totally 29 3-item core herb groups were obtained, and the top three actions of relieving exterior syndrome, invigoration and calming wind. With support values>80 (frequency), the main effect was relieving exterior syndrome. CVP of selected prescriptions was calculated and categorized by dynasties. The top 10 books with largest reappearing rate and their high frequent herbs, and the top 10 books with largest proportion of all selected prescriptions were obtained. Conclusion The basic law of prescriptions for treating stroke in different medical books was obtained. The dynasty distribution of CVP indicates that the research of Chinese medicine for stroke increasingly getting systematic and becoming more accurate in choosing herbs. The principles of choosing herbs to combine a prescription for treating stroke are similar among the collections of selected medical books.