Abstract: A constrained total least squares (CTLS) algorithm based on the bistatic range (BR) and direct range (DR) measurements for the multistatic passive radar localization problem with transmitter and receiver errors. An auxiliary variable is selected to linearize the nonlinear BR equations, and the algebraic characteristics of pseudo linear localization equations of BR is introduced to the DR measurement. Then a constrained total least squares (CTLS) model is constructed, which is solved by Newton's method. The Cramer-Rao Lower Bound (CRLB) and the theoretical error of the algorithm are then derived and analyzed. Simulation experiments show that the estimated error of the proposed CTLS algorithm can reach CRLB under moderate level of noises. Its position error is lower than that of the two-step weighted least squares under high level of noises, which means a better noise adaptation level.