IR3535, KBR3023, para-menthane-3,8-diol (PMD), and deet were evaluated in controlled studies with human subjects (n = 5) for repellency to black salt marsh mosquitoes (Ochlerotatus taeniorhynchus Wiedemann), in the Everglades National Park, FL. In tests of 6-h duration, with an average mosquito biting pressure on exposed forearm skin of 19.5 (±13.7) bites per minute, the mean percent repellencies (SE) for IR3535, KBR3023, PMD, and deet was 88.6 (3.2), 97.5 (1.7), 89.2 (2.9), and 94.8 (2.5), respectively. Mean complete protection times (SE) for IR3535, KBR3023, PMD, and deet were 3.0 (1.0), 5.4 (0.6), 3.8 (1.4), and 5.6 (0.5) h, respectively. Untreated (ethanol) controls provided 0% repellency. When mosquito biting rates on the untreated forearm skin of repellent-treated subjects were compared with biting rates on the forearm skin of control subjects, the former were 23%–40% lower early in tests and as much as 22% higher late in tests. These differences cast doubt on the technical merit of test designs comprising evaluation of more than one repellent at a time on the same human subject while underscoring the importance of untreated subjects as negative controls in field repellent bioassays.