Abstract: Light attenuation in thick biological tissues, caused by a combination of absorption and scattering, limits the penetration depth in multiphoton microscopy (MPM). Both tissue scattering and absorption are dependent on wavelengths, which makes it essential to choose the excitation wavelength with minimum attenuation for deep imaging. Although theoretical models have been established to predict the wavelength dependence of light attenuation in brain tissues, the accuracy of these models in experimental settings needs to be verified. Furthermore, the water absorption contribution to the tissue attenuation, especially at 1450 nm where strong water absorption is predicted to be the dominant contributor in light attenuation, has not been confirmed. Here we performed a systematic study of in vivo three-photon imaging at different excitation wavelengths, 1300 nm, 1450 nm, 1500 nm, 1550 nm, and 1700 nm, and quantified the tissue attenuation by calculating the effective attenuation length at each wavelength. The experimental data show that the effective attenuation length at 1450 nm is significantly shorter than that at 1300 nm or 1700 nm. Our results provide unequivocal validation of the theoretical estimations based on water absorption and tissue scattering in predicting the effective attenuation lengths for long wavelength in vivo imaging.