A strategy for the optimal operation of multiple-fraction batch distillation processes with waste cut recycling is presented. The processes are modeled by differential algebraic equations formulated in the gPROMS modeling language. A cost function based on economic criteria is developed. Recycling of the waste cuts is formulated as a cyclically repeated batch distillation process, where the waste cuts are recycled at intermediate points in time. The resulting multipoint boundary value optimization problem is efficiently solved by using the dynamic optimization package MUSCOD-II. A comparison of optimization results for the separation of a ternary mixture shows that considerable gains in profits can be obtained by a complete recycling of the waste cuts. Computation times are in the range of minutes on a standard workstation.