Abstract—A four-dimensional method of optimization of spatial-kinematic models of subsystems of objects of the Galaxy based on the principle of maximum likelihood has been proposed, taking into account the measurement and natural (dynamic) uncertainty of 3D velocities and random errors of heliocentric distances (in this case, trigonometric parallaxes). The method has been tested on masers in the high-mass star-forming regions (HMSFRs). Based on the data on these objects, new estimates of the fundamental parameters of the Galaxy were obtained, free from systematic biases due to parallax errors, in particular, the distance from the Sun to the center of the Galaxy R0 = 7.88 ± 0.12 kpc, the angular azimuthal velocity of the Sun = 30.40 ± 0.20 km/s/kpc, the linear azimuthal velocity of the Sun = 239.6 ± 4.0 km/s/kpc.