The goal of this work is twofold: to predict the phase equilibria behavior of simplified surfactant models and to predict the population of aggregates as a function of pressure. We compare Monte Carlo simulation results of these systems with predictions from a modified version of the statistical associating fluid theory (soft-SAFT). Surfactant-like molecules are modeled as Lennard-Jones chains of tangent segments with one or two association sites. We study the influence of the number and location of the association sites on the thermodynamic properties and fraction of nonbonded molecules in all cases. The influence of the chain length is also investigated for a particular location of the sites. Results are compared with NPT Monte Carlo simulations to test the accuracy of the theory, and to study the molecular configurations of the system. Soft-SAFT is able to quantitatively predict the MC PVT results, independently of the location of the association sites. The theory is also able to capture the qualitative trend of the population of aggregates with pressure. Quantitative agreement is only obtained for specific locations of the sites. (C) 2004 American Institute of Physics.