The accuracy and completeness of KC60 coding is important to ensure that accurate local, regional and national data on sexually transmitted infections (STIs) are produced. Following alterations to the codes in 1995 a review of the process and accuracy in one department of genitourinary medicine was undertaken. Over a 6-month period 239 episodes and 407 diagnoses were reviewed in the notes of 172 patients. An error was found in 17.9% of diagnoses, and 32.6% of episodes had at least one error. Only one episode had been diagnosed but not coded. Subsequently the procedure for diagnosing episodes was altered and a subsequent review of 122 episodes and 239 diagnoses in the notes of 94 patients identified only 4 errors.