The circadian and neurobehavioral effects of light are primarily mediated by a retinal ganglion cell photoreceptor in the mammalian eye containing the photopigment melanopsin. Nine action spectrum studies using rodents, monkeys, and humans for these responses indicate peak sensitivities in the blue region of the visible spectrum ranging from 459 to 484 nm, with some disagreement in short-wavelength sensitivity of the spectrum. The aim of this work was to quantify the sensitivity of human volunteers to monochromatic 420-nm light for plasma melatonin suppression. Adult female (n=14) and male (n=12) subjects participated in 2 studies, each employing a within-subjects design. In a fluence-response study, subjects (n=8) were tested with 8 light irradiances at 420 nm ranging over a 4-log unit photon density range of 10(10) to 10(14) photons/cm(2)/sec and 1 dark exposure control night. In the other study, subjects (n=18) completed an experiment comparing melatonin suppression with equal photon doses (1.21 x 10(13) photons/cm(2)/sec) of 420 nm and 460 nm monochromatic light and a dark exposure control night. The first study demonstrated a clear fluence-response relationship between 420-nm light and melatonin suppression (p<0.001) with a half-saturation constant of 2.74 x 10(11) photons/cm(2)/sec. The second study showed that 460-nm light is significantly stronger than 420-nm light for suppressing melatonin (p<0.04). Together, the results clarify the visible short-wavelength sensitivity of the human melatonin suppression action spectrum. This basic physiological finding may be useful for optimizing lighting for therapeutic and other applications.