Background: There is a relative lack of epidemiological data on cancer in Turkey, which is a large country with a population of 71 million. The first population-based registry in the country is Izmir Cancer Registry (ICR) which was not established until 1992. The present study, aiming to address the gap in this kind of epidemiological data for this part of the world, reports the incidence of cancers of the genitourinary tract in Izmir province over a ten year period. Methods: Cancer incidence data for 1993-2002 was obtained from the ICR database, which employs a population based registry system, and actively collects data by following international registration rules. Annual crude and age standardized incidence rates were calculated for the whole period and also for ear lier and later periods. Results: The age-standardized incidence rate (world population) for all sites was 198.3 per 100,000 for males and 116.4 per 100,000 for females. The most common primary sites for men were lung (35.6%), bladder (7.8%), colon and rectum (6.1%), larynx (5.7%) and prostate (5.4%). For women, the principal cancers were breast (28.7%), colon and rectum (7.2%), corpus uteri (5.3%), cervix uteri (4.8%) and lung (4.7%). Urogenital cancers accounted for 11.2% of all new cancer cases for the 1993-2002 period in Izmir. Of the total, 89.6% were observed in males and 10.4% were diagnosed in females. Carcinoma of the bladder was the most common among the urogenital cancers in Izmir province (Age standardized incidence rate, world standard population 17.1 per 100 000). Conclusions: Bladder cancer incidences were quite high, especially for men, and appear to be increasing. Prostate cancer has lower incidence rates as compared to western countries, but the trend is for rise. Although there might be an underestimate of incidences, owing to an inability to use data from death certificates, the overall profile is an accurate reflection of incidence in this region of Turkey and provides adequate information for planning strategies for cancer control.