These notes grew out of a reading course I gave out of Cornuejol's book (Combinatorics: Packing and Covering) at Simon Fraser and were influenced by a course given by Seymour many years earlier at Princeton. My development of the subject differs markedly from that found in Cornuejols. One organizational difference is that I have attempted to treat packing and covering in parallel to emphasize the similarities. So, for instance, I introduce the notion of antiblocker alongside that of blocker and give it equal play. Probably the most significant difference is that I postpone linear programming until such concepts as MFMC, Ideal, and Perfect have already been developed (I introduce blocking and antiblocking polyhedra to prove results such as C ideal iff b(C) ideal). This appeals to me, and I think it has some pedagogical advantages, but has the disadvantage that the relationship between MFMC and Ideal is derived only later on as a theorem.