Photobiomodulation (PBM) refers to a non-destructive, non-thermal application of near infrared or infrared light (lasers or LEDs) on human tissues for medical benefits, such as for wound healing and pain reduction. The mechanism of PBM was understood as the absorption of photon energy by cytochrome-c-oxidase (CCO), the last enzyme in the mitochondrial respiratory chain that catalyzes the reduction of oxygen for energy metabolism. However, it is unclear which wavelengths would be optimal for PBM applications due to lack of systematic studies on wavelength dependence of PBM-induced effects. In this study, we wish to quantify and compare hemodynamic and metabolic benefits of PBM on human tissue with several central wavelengths. We employed 4 LEDs, whose spectral ranges were commonly used in PBM: 760 nm (FWHM=24nm), 810 nm (FWHM=30nm), 970 nm (FWHM=46nm), and 1070 nm (FWHM=55nm). The power densities of these LEDs were measured and calibrated by a spectrometer and power meter. We applied 5-minute PBM using each of these LEDs in vivo on the right forearm of 15 human participants. Corresponding placebo experiments were also carried out for rigorous comparison. Broadband near infrared spectroscopy (740-900 nm) was employed near the PBM site to quantify hemodynamic and metabolic responses during the PBM/placebo and a 3-minute recovery period. A multi-linear regression analysis based on the modified Beer-Lambert law was performed to estimate changes of oxy-hemoglobin, deoxy-hemoglobin, and cytochrome-c-oxidase concentration. Different patterns of hemodynamic and metabolic responses were observed for these 4 LED wavelengths, showing quantitative wavelength-dependent PBM effects.