BACKGROUND: The aim of the study was to provide data on the relative frequency of reported symptoms in travelers using chloroquine, chloroquine plus proguanil, and mefloquine. METHOD: The study was an open, nonrandomized study recording self-reported events in travelers recruited consecutively from two travel clinics in Copenhagen, Denmark. The main outcome measures were the relative proportion of travelers reporting particular symptoms in the three prophylaxis groups, compliance, hospitalization and premature termination of the travel. RESULTS: From May 1996 to April 1998 5, 446 travelers were included and 4,158 questionnaires (76.3%) returned. Compliance was significantly better in mefloquine users with 83.3% of short term travelers compared to 76.3% in chloroquine plus proguanil users. Also, 84.8%, 59.3% and 69.5% using chloroquine, chloroquine plus proguanil, and mefloquine respectively reported no symptoms and 0.6%, 1.1% and 2.8% reported "unacceptable" symptoms. Compared to chloroquine, mefloquine users had a significantly higher risk of reporting depression, RR 5.06 (95% CI 2.71 - 9.45), "strange thoughts," RR 6.36 (95% CI 2.52 - 16.05) and altered spatial perception, RR 3.00 (95% CI 1.41 - 6.41). CONCLUSION: Overall mefloquine is tolerated at least as well as chloroquine plus proguanil and shows better compliance, however, symptoms related to the central nervous system are more prevalent in mefloquine users and when symptoms develop, they are perceived as more severe.