Two cross-sectional surveys were undertaken, from December 1982 to August 1983 and from November 1990 to January 1991, to estimate the prevalence rates of genital ulcer disease (GUD) in all patients presenting with a new sexually transmitted disease (STD) complaint to the STD clinic at the Comprehensive Health Centre in Kingston, Jamaica. Diagnosis of syphilis and human immunodeficiency virus (HIV) infection was based on results of laboratory tests, but diagnosis of the other STDs was based on clinical features. Data from these two surveys were compared, and reported national annual incidence data for GUD reviewed. In 1982/83 6.8% of 23,050 patients had GUD, men (9.3%) more often than women (4.2%; p < 0.001). In 1990/91 the prevalence rate was 12.8%, with increased rates for both men (18.2%) and women (6.8%; p < 0.001). In patients with GUD, a clinical diagnosis of genital herpes was made, in 1982/83 and 1990/91, respectively, in 16.8% and 7.8% of the patients; syphilis, in 12.9% and 18.8%; chancroid, in 12.4% and 13.3%; viral warts, in 5.7% and 6.3%; lymphogranuloma venereum, in 4.1% and 3.9%; and granuloma inguinale, in 3.6% and 2.3%. In men the rate for syphilis was 19% in 1990/91 and 8% in 1982/83 (p = 0.001); and for genital herpes it was 7% in 1990/91 and 17% in 1982/83 (p = 0.025). These reversals were attributed to intense media coverage of herpes in 1982/83. There was no difference in prevalence rates between the two surveys for these diseases in women, or for lymphogranuloma venereum, granuloma inguinale and genital warts in men and women. A clinical diagnosis could not be made in 44.4% of cases in 1982/83 (particularly in men), and in 47.6% of cases in 1990/91. GUDs facilitate transmission and adversely affect the prognosis of HIV. The increase in their prevalence has implications for the evolution of the local HIV epidemic, and should be addressed effectively by strengthening the STD/HIV control programme.