An estimated 50 million people each year from industrialized countries visit tropical areas: 3% to 11% of these travelers report a febrile illness on their return. We conducted a 5-year prospective observational study on the causes of fever in patients admitted to a university teaching hospital after returning from the tropics. We enrolled in this study all consecutive patients admitted to the Division of Infectious Diseases of the University of Milan, Italy, between January 1997 and December 2001 presenting with fever (oral temperature > or =37.5 degrees C) and a history of travel to a tropical country in the previous 6 months. Seven percent (147/2,074) of all hospital admissions in the study period were due to fever in travelers and migrants returning from the tropics. Malaria accounted for 47.6 % of all admissions (70/147), followed by presumed self-limiting viral infections (12%). Pretravel screening and vaccination strategies could have prevented a considerable number of hospitalizations (e.g., hepatitis A and typhoid fever). The most useful investigations were blood examination and PCR for malaria, which gave positive results in 65% of cases in which they were performed. During a 5-year period, the number of patients returning from tropical areas who were admitted with fever to a university hospital in northern Italy remained stable; malaria was the most frequent diagnosis, and should be considered in any febrile patient returning from the tropics. With the exception of hepatitis A and dengue fever infections, in a real-world setting serology is of modest utility and is probably overused.