Acoustic tomography of the atmospheric surface layer is based on measurements of travel times of sound propagation among different pairs of sources and receivers usually located several meters above the ground on a horizontal scale of about 100 m. The measured travel times are used as input data in an inverse algorithm for reconstruction of temperature and wind velocity fields. Improved knowledge of these fields is important in boundary layer meteorology, theories of turbulence, and studies of electromagnetic and acoustic wave propagation in the atmosphere. In this paper, a short overview and current status of acoustic travel-time tomography of the atmosphere are presented. A brief description of a 3D array for acoustic tomography of the atmosphere which is being built at the Boulder Atmospheric Observatory is given. Furthermore, different inverse algorithms for reconstruction of temperature and velocity fields are discussed, including stochastic inversion and a recently developed time-dependent stochastic inversion. The latter inverse algorithm was used to reconstruct temperature and wind velocity fields in acoustic tomography experiments. Examples of the reconstructed fields are presented and discussed.