Workload　programming　in　discrete　manufacturing　dynamic　systems　is　allocating　suitable　workloads　of　every　production　periods　according　to　the　needs　of　products,　which　would　minimize　the　total　cost　of　both　workload　and　semi-product　stock　in　whole　production　process　under　some　constraint　conditions　of　systems.　In　this　paper,　a　dynamic　equation　described　the　relations　between　stocks　and　workloads　is　established　by　means　of　petri　net.　A　total　optimization　approach　allocating　workloads　of　all　periods　is　provided　using　linear　integral　programming　and　dynamic　programming.　An　example　is　given　to　verify　this　approach.