A computer program package for use with the southern pine seed orchard Inventory-Monitoring System (IMS) was developed and tested using 1980 and 1981 cone crop data collected from Weyerhaeuser's loblolly pine seed orchard at Magnolia, Arkansas. The Computerized IMS, written in Fortran, is accompanied by a user's guide containing data entry instructions. The IMS package includes one predictive program and two programs designed to interpret actual harvest data and make it available for the computation of predictions for future crops. The predictive program utilizes strobilus survival data from sample trees and clonal estimates for cone analysis-derived variables to compute clonal cone and seed yield predictions. Cone yield predictions calculated by the Computerized IMS for the 1980 and 1981 crops indicated that it has the potential for producing useful production estimates. Investigations of variance within the seed orchard having an impact on IMS predictions centered on clonal, age-class (within-clone), annual, orchard location and with in-crown effects on cone analysis-derived variables, cone efficiency and flowers-per-tree. Analysis of clonal variance indicated that seed potential and germination efficiency were under strong genetic control compared to the other characters. Annual clonal stability was high enough for clonal seed potential and seed efficiency to make useful regression models possible for estimating these variables without complete cone analysis. Clonal cone efficiencies displayed low annual stability, discouraging the use of stratified clones to estimate productivity, since clones could change productivity classes from year to year. Within-crown variation was not well-defined for most variables due to clone-crown sector interaction. However, cones were concentrated in the middle one-third of the crown and empty seed percentages were found to be highest in the north crown quadrant. The study indicated that clone-by-clone monitoring of seed orchard production, while the most expensive option, will produce the most reliable and useful yield predictions. / Ph. D.