This research develops and tests a methodology for improving quality on a continuous process line. A nonwovens line was chosen as the experimental environment to test this methodology. First the process variables were analyzed and classified and a screening experiment was run to determine which variables had a significant effect on the quality measures--air permeability, caliper, basis weight, and appearance--of the filter media. Then a second experiment was run to study fewer variables in more detail. The data were interpreted with ANOVA and regression models. This research makes three general contributions: a methodology for improving quality on a continuous process line, the cause-and-effect relationships between some of the independent and dependent variables, and the second-order and third-order effects show the complexity of the research environment.