This research will look at a printing operation that prints labels for paint cans. This particular printing operation is a multi-product, single-machine system with stochastic demands that uses a printing press to print the labels on large sheets that may contain one label type (dedicated sheets) or many different label types (mixed sheets). The objective of this thesis is to determine whether to set up the press to print mixed sheets or dedicated sheets in order to minimize the total cost of the current system given three scenarios: one where all label types have approximately the same monthly demand, one where 20% of the label types have very large demand and the remaining 80% have small demand, and one where half of the label types have moderately high demand and the other half have moderately low demand. Three simulation models, one for each scenario, are developed to test this objective and to see which scenarios prefer mixed sheets and which prefer dedicated sheets. The decision logic to choose the sheet to print is based on summing the differences of the inventory levels and the reorder points for each label type on the sheet, and choosing the sheet with the minimum value of this equation. The results of the simulations show that, according to this decision logic, all scenarios choose dedicated to print dedicated sheets.