Technical Abstract: Many landowners in the United States have little knowledge of the potential economics returns from agroforestry practices. Economic simulators for temperate agroforestry practices have been generated; yet, there are few data sets on yields of timber and other products to run, validate and refine such models. The objectives of this study were to characterize variations in nuts yields among eastern black walnut (Juglans nigra L.) trees and apply this information to the development of predictive equations between tree diameter at breast height (DBH) and nut yields. Three data sets were analyzed that included results from Tennessee, Chetopa, Kansas and Mt Vernon, Missouri. Tree-to-tree variation in nut yields was high within each data set, with coefficient of variations for nut yields, typically exceeding 50%. Averaging nut yields over several consecutive years reduced coefficients of variation. Nearly half of the high nut producing trees exhibited biennial pattern of alternate bearing of nuts. Trees with low average nut yields had either sporadic or irregular patterns of nut bearing. The regression coefficients for equations relating stem diameter and nut yields varied considerably. Averaging nut yields over consecutive years, and averaging stem diameter and nut yields over a number of trees increased regression coefficients of such equations. These results indicate that predicting nut yields of a tree stand over a several year-period will be easier than predicting yields for a specific tree in a specific year.