The invention relates to application of set pair analysis in representative prescription use law research. The application includes the following steps that: step S1, the name of a
disease is selected, representative prescriptions which are reported in literature and are used for treating the
disease are screened out; step S2, a four-element
connection number method is adopted: the clinical efficacies of the representative prescriptions screened out in the step S1, namely "cured", "significant effect", "improved" and " inefficient " are a four-element
connection number which is expressed as anequation that Mu=a+bi+cj+dk in set pair analysis; step S3, a distance to a perfect
efficacy is calculated: a distance between the four-element
connection number of each prescription which is expressed as an equation that Mut=at+bti+ctj+dtk and the four-element connection number of the ideal perfect
efficacy which is expressed as an equation that Mu=1+0i+0j+0k; and step 4, medication laws are searched: the medication laws of some prescriptions in all the representative prescriptions are examined, wherein the
system efficacies of the certain prescriptions rank first three places, and the prescriptions are clustered according to medication frequencies; and step 5, comparative discussion is carried out. With the application of the invention adopted, general medication laws in different
syndrome differentiation and treatment prescriptions are explored; the efficacies of all the representative prescriptions are sequenced according to the merits and demerits of the representative prescriptions on the basis of total
efficacy rates; and therefore, clinical efficacies can be improved.