A large collection of radiometric signature data on various military targets was utilized to develop a simplified multiple linear regression technique for signature extrapolation. Environmental considerations were limited to air temperature, solar radiation, sky temperature, and wind speed. Targets were broken into about ten distinct classes of materials for predictive purposes. Results are presented for a number of the classes of materials found on typical targets. Data are compared with regression predicted values to evaluate the goodness of fit. Extrapolated signature values for a tank target are compared with predictions of two of the more comprehensive thermal models.