Multispectral LiDAR systems have been proved to have potential for retrieving forest structural and physiological parameters from waveforms of various wavelengths. However, multiple scattering occurring in complex forest canopy directly affects the waveform shape. This study combined a leaf optical model, a forest structure model, and a LiDAR process model to simulate multispectral LiDAR waveforms in single and multiple scattering cases and then assessed the effect of multiple scattering on forest physiological indicators. Results showed that 1) multiple scattering increases the waveform intensity, especially at near-infrared wavelength; 2) multiple scattering greatly affects the retrieval of physiological parameters of the understory; 3) when leaf chlorophyll content is low, forest physiological indicator was relative-seriously overestimated due to the multiple scattering effect. These findings may shed some light on our understanding of multiple scattering mechanism and accurately retrieving forest physiological parameters from multi-spectral LiDAR waveforms.