We analyze the complexity of reasoning in EL with defeasible inclusions, and extend previous results by tightening lower and upper complexity bounds and by relaxing some syntactic restrictions. We further extend the old framework by supporting arbitrary priority relations over defeasible inclusions.