It seems odd that male suicide rates have decreased in Portugal over the last few years, whereas the opposite has happened in most other countries. In order to investigate the reasons for such a decrease, suicide rates were compared with controversial cases (ICD-9 category E 980-989) and with homicide rates. The results show that, since 1980 and coinciding with the use of ICD-9, controversial cases have increased 12 and 21 times among males and females respectively. The profile of controversial cases is similar to that of suicide according to sex, age, marital status and season. Rates for homicide have remained steady and have a distinct profile. It is concluded that since 1980 there has been a significant underreporting of suicides in Portugal, which have been registered as controversial cases. The difficulty of investigating and preventing suicide with such statistical data is stressed; an improvement in suicide reporting and registering is urgently required.