Received XX Month XXXX; revised XX Month, XXXX; accepted XX Month XXXX; posted XX Month XXXX (Doc. ID XXXXX); published XX Month XXXX We propose a method to reconstruct the two-dimensional (2D) velocity distribution of flow fields by laser absorption spectroscopy tomography. A mathematic model is established to reveal the dependence of spectral absorbance on line-of-sight velocity distribution. Then, with multiple laser beams from different angular views covering the region of interest, a nonlinear equation set of 2D velocity distribution is established according to the model. The integrated absorbance coefficient distribution is reconstructed using the Landweber iteration algorithm and substituted into the nonlinear equation set for further simplification. Finally, the velocity distribution is reconstructed by solving the simplified equation group via the interior-point algorithm. The proposed method is validated numerically by reconstructing the velocity distribution of water molecules, as calculated by Computational Fluid Dynamics, over a cross-section of a double-mode scramjet combustor. The method does not require adding extra tracer particles and avoids issues arising from the short lifetime of molecular tags. It is suitable for diagnosis of high-speed flow fields.