Analysis of existing gaseous emissions data from gas turbines has shown that a substantial improvement in data correlation is possible. These correlations serve to increase confidence in the data, identify out-of-control points, and facilitate comparisons between engine models. Furthermore, the NOx correlation implies that a simplified primary zone mathematical model may be used to formulate trends or influence coefficients, which may be used to adjust experimental data to a common base for comparison to a specification. Finally, the presence of water vapor in the combustion air is seen to have a noticeable effect on NOx emissions. This effect should be accounted for in comparing experimental results.