Statistical process control (SPC) charts have become widely implemented tools for quality monitoring and assurance in healthcare settings across the United States. SPC methods have been successfully used in industrial settings to track the quality of products manufactured by machines and to detect deviations from acceptable Levels of product quality. However, problems may arise when SPC methods are used to evaluate human behavior. Specifically, when human behavior is tracked over time, the data stream generated usually exhibits periodicity and gradualism with respect to behavioral changes over time. These tendencies can be quantified and are recognized in the statistical field as autocorrelation. When autocorrelation is present, conventional SPC methods too often identify events as "unusuaL" when they really should be understood as products of random fluctuation. This article discusses the concept of autocorrelation and demonstrates the negative impact of autocorrelation on traditional SPC methods, with a specific focus on the use of SPC charts to detect unusual events.