In today's global market, quality improvement has become an essential element of remaining competitive. Given a stable manufacturing process, there are two competing strategies for improving it. The first, a conventional approach, relies on a “one factor at a time” strategy, usually requires added costs, and is often limited in its success. The second approach relies on methods grouped under the name “Statistical Problem Solving” (SPS) and simultaneously exploits statistical science, teamwork, existing process knowledge, and execution strategies. Problems are solved and processes improved by reducing statistical variation at virtually zero cost. This paper reviews the conventional problem-solving approach with some of its shortcomings, then systematically presents the SPS strategy.