International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) has been conducting a prospective evaluation period to validate the criteria for waiving some carcinogenicity studies in rats. Before the waiving strategy is practiced in ICH, it is crucial to elucidate whether non-neoplastic lesions are found only in 2-year rat carcinogenicity studies. To confirm possible importance of 2-year bioassays for evaluating chronic toxicity but not carcinogenicity, we retrospectively surveyed 59 pharmaceuticals approved by the Ministry of Health, Labour and Welfare (MHLW) from 2007 to 2010 in Japan for non-neoplastic lesions observed in carcinogenicity studies. Non-neoplastic histopathological lesions observed only in 2-year carcinogenicity studies but not in 6-month chronic toxicity studies using rats were compared with clinical adverse drug reactions (ADRs). Thirteen non-neoplastic lesions that may correlate with clinical ADRs were classified into three categories: Category 1, lesions not predictable from other nonclinical data except those from 2-year rat carcinogenicity studies; Category 2, lesions predictable mainly from chronic toxicity studies; Category 3, lesions predictable mainly from pharmacological actions. In the present survey, non-neoplastic lesions only found in 2-year rat carcinogenicity studies were neither significant in terms of frequency and severity nor useful for clinical risk management.