Dengue often presents with non-specific clinical signs, and given the current paucity of accurate, rapid diagnostic laboratory tests, identifying easily obtainable bedside markers of dengue remains a priority. Previous studies in febrile Asian children have suggested that the combination of a positive tourniquet test (TT) and leucopenia can distinguish dengue from other febrile illnesses, but little data exists on the usefulness of these tests in adults or in the Americas. We evaluated the diagnostic accuracy of the TT and leucopenia (white blood cell count <5000/mm(3)) in identifying dengue as part of an acute febrile illness (AFI) surveillance study conducted in the Emergency Department of Saint Luke's Hospital in Ponce, Puerto Rico. From September to December 2009, 284 patients presenting to the ED with fever for 2-7 days and no identified source were enrolled. Participants were tested for influenza, dengue, leptospirosis and enteroviruses. Thirty-three (12%) patients were confirmed as having dengue; 2 had dengue co-infection with influenza and leptospirosis, respectively. An infectious etiology was determined for 141 others (136 influenza, 3 enterovirus, 2 urinary tract infections), and 110 patients had no infectious etiology identified. Fifty-two percent of laboratory-positive dengue cases had a positive TT versus 18% of patients without dengue (P<0.001), 87% of dengue cases compared to 28% of non-dengue cases had leucopenia (P<0.001). The presence of either a positive TT or leucopenia correctly identified 94% of dengue patients. The specificity and positive predictive values of these tests was significantly higher in the subset of patients without pandemic influenza A H1N1, suggesting improved discriminatory performance of these tests in the absence of concurrent dengue and influenza outbreaks. However, even during simultaneous AFI outbreaks, the absence of leucopenia combined with a negative tourniquet test may be useful to rule out dengue.