Four rhesus monkeys were tested for preferences within the wavelength and line-tilt dimensions. In the case of wavelength, the response panel was back-illuminated by light of one of the following wavelengths, presented in a random manner: 470, 525, 580, and 635 nm. Similarly, the line-tilt dimension was studied, by presenting a 5 cm by 0.3 cm black bar tilted at 0, 30, 60, or 90 degrees. No preferences were found within this latter dimension; in contrast, marked wavelength preferences existed, the order of preference being 470 (most preferred), 525, 580, anid 635 nm (least preferred). These response preferences were resistant to behavioral manipulation; the number of responses to blue and to red in extinction was about equal when red was used as the training stimulus, but vastly different following training on blue. These results indicate that such response preferences must be taken into account in the design of a wide variety of experiments.