In this paper, we present a novel methodology for processor evaluation in an embedded systems design environment. This evaluation can help in either selecting a suitable processor core or in evaluating changes to an ASIP. The processor evaluation is carried out in two stages. First, an architecture independent stage in which processors are rejected based on key application parameters and secondly, an architecture dependent stage in which performance is estimated on selected processors. The contribution of our work includes identification of application parameters which can influence processor selection, a mechanism to capture widely varying processor architectures and an instruction constrained scheduler. Initial experimental results suggest the potential of this approach.