BACKGROUND: Most antimalarial agents used by travelers act on the parasite's blood stage and therefore do not prevent late-onset illness, particularly that due to species that cause relapsing malaria. We examined the magnitude of this problem among Israeli and American travelers. METHODS: We examined malaria surveillance data from Israel and the United States to determine the traveler's destination, the infecting species, the type of chemoprophylaxis used, and the incubation period. RESULTS: In Israel, from 1994 through 1999, there were 300 cases of malaria among returning travelers in which one species of plasmodium could be identified. In 134 of these cases (44. 7 percent), the illness developed more than two months after the traveler's return; nearly all of these cases were due to infection with Plasmodium vivax or P. ovale. In 108 of the 134 cases (80.6 percent), the patient had used an antimalarial regimen according to national guidelines. In the United States, from 1992 through 1998, there were 2822 cases of malaria among travelers in which the cause could be evaluated. Late illness developed in 987 (35.0 percent) of these travelers. The infection was due to P. vivax in 811 travelers, P. ovale in 66, P. falciparum in 59, and P. malariae in 51; 614 (62.2 percent) of those with late-onset illness had appropriately taken an effective antimalarial agent. CONCLUSIONS: In more than one third of malaria-infected travelers, the illness developed more than two months after their return. Most of these late-onset illnesses are not prevented by the commonly used and effective blood schizonticides. Agents that act on the liver phase of malaria parasites are needed for more effective prevention of malaria in travelers.