This paper describes the progress of and the challenges facing SSA’s Cohorts model.  Like other open models (LifePaths, SSASIM), the Cohorts model can simulate individual birth cohorts without needing, as in closed models, to simulate the mating and birth processes that generate the cohorts.  By simulating successive cohorts the model builds up year by year cross-sectional aggregates, which can be calibrated against historical or projected aggregates.  The Cohorts model differs from other microsimulation models in using sampled lifetime error configurations for the individual earnings histories, rather than generating the error configurations from scratch from a modeled and estimated error process.  Most of the paper will focus on the advantages and pitfalls of this last aspect.