Data from Georgia timber product output studies were used to develop models that spatially allocate roundwood receipts data from primary wood-using mills. Mill receipts data were converted to a cumulative frequency based on a distance from the mill to a county center. Distances were calculated as either straight line or shortest road distance, and county centers were determined as geographic centers or weighted by forest mass. Logistic and Gompertz nonlinear asymptotic model forms were used to model cumulative frequency of annual receipts within a repeated measures framework. It was determined that the number of mill employees and the shortest road distance were significant factors. The Gompertz model form predicted an unbiased cumulative frequency for the 2003 data using several spatial covariance types in the repeated measures analysis. These methods demonstrate the potential to spatially allocate a mill's total receipts to its surrounding counties in the absence of county-specific recording.