Similarity in drug names is one of the causes of errors in prescription and medicine dispensation. Therefore, drug names used by pharmaceutical companies should be as dissimilar as possible. However, they cannot avoid using similar names for generic drugs which use common names. Therefore, we divided drug names that can be expressed in katakana in Japan into two groups: Group A, which does not include generics, and Group B, which all brand-name drugs that can be switched to generics are replaced by generics, and we comprehensively calculated and compared the name similarity. The name similarities were scored using m2-vwhtfrag. If the score exceeded 0.46, the names were considered similar. The ratio of combinations with similar names by dosage form were：oral, 2.78％ in group A and 2.48％ in group B, topical, 3.06％ in group A and 2.96％ in group B, and injection, 2.85％ in group A and 2.63％ in group B. The percentage of similarity of drug names per dosage form decreased when the all brand-name drugs was replaced by generics.