During August and September 1951, forest canopy mosquito collections were made at two localities in Costa Rica where sylvan yellow fever was active. Haemagogus spegazzinii falco, a species which has been found naturally infected with yellow fever on a number of occasions in Colombia, was the commonest species of the genus present at Wauchope and was at least as abundant at San Gerardo as it had been at localities where yellow fever appeared in Panama in the two years previous. The conservative course would be to take this association at face value and consider this species the probable vector in this outbreak.
Haemagogus spegazzinii falco is ordinarily an inhabitant of the upper stories of primary tropical rain forest, but at Wauchope we found it commonly at ground level in cacao plantation in which the original forest cover had been thinned to provide the partial shade necessary for cacao trees. In these circumstances the canopy habitat is simulated at ground level and the stratification of spegazzinii falco is less pronounced. Such a situation provides conditions under which there is danger of sylvan yellow fever transmission on the ground.
Other Haemagogus present were H. equinus and H. lucifer-iridicolor. The former is known to be capable of transmitting virus in the laboratory but has never been found infected with certainty in nature. Nothing is known of the vector status of the latter two species. As the numbers of H. spegazzinii falco present were ample to explain the transmission of yellow fever in the area, these species need not be considered in the present circumstance.
The commonest sabethine mosquito present was Sabethes chloropterus. Its status as a vector is unknown, but on epidemiological grounds we considered it suspect in Panama, and note with interest its present association with yellow fever.