We explore three recursive Bayesian algorithms for the passive location of a stationary ground-based emitter using bearing measurements obtained by a flying platform. The passive location of emitters is a frequent requirement in rescue missions. The algorithms that we consider are a particle filter, the unscented Kalman filter (UKF) and the extended Kalman filter (EKF). They require increasingly greater simplification to the models in the passive location problem. The particle filter is a straightforward formulation based on rejection sampling, requiring the measurements to be conditionally independent and the likelihood to be bounded with known bound. Apart from these requirements, which are satisfied for the problem, no further simplification is needed. The UKF preserves the nonlinearities in the models but approximates the posterior distribution at each time step by a Gaussian distribution. The EKF also assumes a Gaussian distribution, but further linearizes the models so that Gaussianity is preserved under the simplified linear models. Our Monte Carlo simulation results show that the recursive particle filter converges more quickly than the UKF and EKF, and performs better in terms of both point estimation and distribution estimation.