From July to December 1998, a hospital- and health center-based surveillance system for dengue was established at selected sites in Nicaragua to better define the epidemiology of this disease. Demographic and clinical information as well as clinical laboratory results were obtained, and virus isolation, reverse transcriptase-polymerase chain reaction, and serologic assays were performed. World Health Organization criteria were used to classify disease severity; however, a number of patients presented with signs of shock in the absence of thrombocytopenia or hemoconcentration. Therefore, a new category was designated as "dengue with signs associated with shock" (DSAS). Of 1,027 patients enrolled in the study, 614 (60%) were laboratory-confirmed as positive cases; of these, 268 (44%) were classified as dengue fever (DF); 267 (43%) as DF with hemorrhagic manifestations (DFHem); 40 (7%) as dengue hemorrhagic fever (DHF); 20 (3%) as dengue shock syndrome (DSS); and 17 (3%) as DSAS. Interestingly, secondary infection was not significantly correlated with DHF/DSS, in contrast to previous studies in Southeast Asia. DEN-3 was responsible for the majority of cases, with a minority due to DEN-2; both serotypes contributed to severe disease. As evidenced by the analysis of this epidemic, the epidemiology of dengue can differ according to geographic region and viral serotype.