A new approach for the estimation of bid-rent functions for location choice is proposed. The method considers that the expected maximum bid in the auction of a good is a latent variable than can be related to observed prices for similar goods. The model generates estimates for the spatial distribution of agents and prices simultaneously that are better than those obtained by estimating a maximum bid and a price model independently. The model is applied and validated for a case study. Results show that the proposed model outperforms other methods for bid rent estimation, especially when detailed data describing the real estate goods an their prices is not available.