Linear regression techniques are used to establish a quantitative description of side forces on bodies at high incidence. A data base is assembled concerning the key side force characteristics of maximum observed side force, angle of occurrence, and minimum angle of attack at which a side force is observed. This information is examined to determine the important trends and a predictive model for side force based on the crossflow analogy is developed to suggest other important variables. A linear regression model for these quantities is developed to include only those variables which are statistically significant. Results indicate that peak side force coefficients are a function of Mach number and only weakly of Reynolds number. Nose fineness is the critical model dimension which suggests that peak side force is a product of the nose flow field. The angle at which peak side forces occur is found to be dependent on model length and Mach number, while the onset angle of attack is a function of model length only.