Systems and methods for optical multi-path interference (MPI) compensation are provided. In an embodiment, a mean MPI signal representing a mean amplitude of the MPI in an input signal is generated and subtracted from a first estimate of transmitted amplitude of the input signal to generate a mean MPI compensated estimate of transmitted amplitude. The mean MPI compensated estimate of transmitted amplitude is sliced to generate a decision of transmitted amplitude of the input signal. The mean MPI signal can be generated using a mean MPI feedback loop or using an iterative feed-forward process. In another embodiment, mean MPI levels corresponding to respective transmitted intensity levels are generated and used to control slice levels of a slicer in order to compensate for MPI.