The structural deficits and problems associated with application of the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) were investigated. For this purpose, mortality statistics in Japan were examined for bias in the proportions of death causes classified using the ICD-10 four-character subcategory system, particularly the large distribution of cases classified under the .9 (unspecified) codes. The results showed that most deaths due to the 3 major causes were coded into the .9 subcategories. For example, 97.6% of the 607,065 deaths between 1995 and 2007 classified as I21 (acute myocardial infarction) were coded as I21.9 (acute myocardial infarction, unspecified). This suggests that the quality of the data recorded on many death certificates is poor. Thus, to ensure that the fourth-digit-level subcategories of the ICD-10 code system are effectively used to represent mortality data and actual causes of death in Japan, future studies should address the following objectives:(1) to minimize the discrepancy between the ICD classification terms and the names of diseases actually used in clinical practice, and (2) to actively raise awareness among physicians about the need to accurately record death causes on death certificates, since these records form the basis for determining mortality statistics in Japan.