The purpose of this study was to develop, implement, and assess an estimation procedure for preventing drug-induced liver disorders. We have built a database for CARPIS (case reports of adverse drug reaction and poisoning information services) since 1987, and the case reports of adverse drug reaction accumulated in the CARPIS database to be total about 11,000. We studied the estimation procedure by evaluating the subjective symptoms, backgrounds and laboratory data of patients in 199 cases cumulated in the CARPIS database. The evaluation scores were created on the basis of the subjective symptoms and backgrounds of the patients. Then, we re-estimated 199 cases with the evaluation scores, among which were 165 cases (82.9%) in 199 cases drug-induced liver disorders. For the rest of 34 cases (17.1%) drug-induced liver disorders could not be estimated from our evaluation scores. These 34 cases were either those which the subjective symptoms were not described in their reports, because the patients were hospitalized for other diseases and drug-induced liver disorders were discovered by a clinical examination, or cases of infants for whom it was difficult to confirm the subjective symptoms. The validity of this evaluation scores were sensitivity = 82.9%, specificity = 91.0% and predictive value of positive test = 94.8%. To apply this evaluation scores onto the clinical practice, we prepared an evaluation form for the subjective symptoms and backgrounds of the patients with drug suspecting-induced liver disorders.