In real-world applications asset managers need to add constraints to their portfolio optimization problem due to several circumstances that may be related with the strategy they want to implement like neutral portfolios, regulatory requirements like limits on minimum and maximum weight per asset, or investor restrictions like not investing in a certain sector or trying to replicate an index performance with a fixed number of assets. These constraints are examples of real feature constraints and most of them can be expressed as constraints or objective functions that can be incorporated directly into the portfolio optimization models that we studied in previous chapters. This chapter explains how most real feature constraints that asset managers face in their asset allocation process can be posed as convex programming or mixed integer programming expressions.