This research produced statistically based, semimechanistic models describing partitioning of Cu and Zn in 40 soils from the United States, Canada, the United Kingdom (UK), The Netherlands, and Chile with widely varying characteristics. Two different types of models were constructed, partitioning models and competitive adsorption models. Multiple linear regression (MLR) was employed to prioritize over 30 different soil characteristics. Multiple linear regression yielded equations predicting the partitioning of Cu and Zn. Equations were also created that estimated the potentially bioavailable fraction of Cu and/or Zn. Data from plant uptake studies (which are reported separately) governed the choice of a suitable chemical soil extraction that estimated bioavailable Cu (0.01 M HCl) and bioavailable Zn (0.01 M CaCl2). Soil pH (1:1 soil:deionized water [DI H2O]) and percent organic matter accounted for approximately 70% of the variability in Cu partitioning and 80% of the variability in bioavailable Cu in the 40 soils studied. For Zn, soil pH alone accounted for roughly 75% of the partitioning variability and 80% of the variability for the estimated bioavailable portion. The results presented here were used in conjunction with results from the plant uptake studies for the creation of models to assess the potential bioavailable metal associated with any given soil from a wide variety of locations.