Dehydrogenation factors, proximate analysis, and ultimate analysis were examined as predictors of the low-temperature carbonization properties of coals varying in rank from lignite through anthracite. All factors studied contributed to reduction in variance of the sample populations investigated. Response surface regression models based on variables from proximate and ultimate analyses of coal were constructed for eight low-temperature carbonization assay product and gas component yields. These models were tested with independent samples that included drill cores, sized fractions, petrographic fractions, and run-of-mine coal. The models developed were used to predict low-temperature carbonization yields for coals from major fields of Wyoming. This information was cast in the form of contour plots describing the potential low-temperature carbonization product yields of Wyoming coals and the properties of the resultant char or coke. The variables used as input to the models may be measured and collected during exploration; thus, the carbonization properties of a seam can be determined in advance of mining. The information presented in this report is useful in that it broadens the decisionmaking base for management when either mine locations or coal conversion plant sites, or both, are being considered.