In this paper we present validation of laser-Doppler spectrum decomposition procedure in estimation of speed distribution of particles. Decomposition method is based on assumption that measured laser-Doppler spectrum can be approximated by linear combination of Doppler shift probability distributions calculated for different speeds of particles and anisotropy of light scattering in the medium. The Doppler shift probability distributions were calculated using Monte-Carlo simulations for Henyey-Greenstein scattering phase function. This decomposition method allows to obtain distribution of speeds of moving particles in the medium, not only average speed as it was possible in laser-Doppler perfusion monitors. Recently we reported that the method was positively verified on spectra generated for different speed distributions using Monte Carlo simulations. In this study we present results of application of the decomposition procedure in analysis of laser-Doppler spectra obtained in physical phantom experiments. A diluted solution of milk was pumped through a tube with different speeds. The dependence of the obtained distributions of speed of moving particles on the speed of flow was observed. Laser-Doppler spectra obtained during in-vivo experiment were also successfully decomposed. A healthy volunteer was investigated and the spectra of laser-Doppler signal during postocclusive hyperemia test were recorded and analyzed. We conclude that the spectrum decomposition procedure can be successfully applied in analysis of the measured laser-Doppler spectra and the amount of information provided by laser-Doppler technique can be significantly increased.