This paper reviews the theoretical background, optimal levels, strengths, weaknesses, and additional considerations of the most frequently used structural equation modeling (SEM) fit statistics in an effort to enable researchers to make better, more informative judgments regarding their models. Fit indices evaluate model fit for the data being examined. Models demonstrate overall fit and the local fit of individual parameters. Some of the commonly used fit indices discussed are: (1) chi squared; (2) goodness of fit and adjusted goodness of fit indices; (3) normed fit and nonnormed fit; (4) comparative fit index; and (5) root mean square error of approximation. Researchers using SEM must determine whether they are interested in testing the null hypothesis, absolute fit, or incremental fit, and they should be aware of the shortcomings of different fit statistics and how the model may lessen the applicability of specific fit statistics. (SLD)