This chapter explains how the Black Litterman model and its variants are used by asset managers to incorporate their views or prior information about the future performance of asset returns or risk factors into their historical estimates of the expected returns vector and covariance matrix of asset returns. This procedure allows to build more robust estimators of the input parameters. These new estimators tend to produce more diversified and robust portfolios than if we only use the sample-based estimates of the parameters as inputs of portfolio optimization models.