Abstract. Hybrid hazelnuts that are predominately a cross between the American hazelnut () and the European hazelnut () are being grown and evaluated as part of an effort to develop a thriving hazelnut industry for the Upper Midwest of the U.S. Along with this plant development effort, researchers are investigating and assessing various harvesting and processing methods and equipment in an effort to create a robust and food-safe production industry. One harvesting alternative is to pick hazelnut clusters off plants before the nuts fully ripen and fall to the ground, an approach that requires greater attention to drying. Whether entire clusters are dried or the nuts are separated from the husks prior to drying is a decision that will be influenced by the drying requirements and potential uses for these hazelnut fractions. To this end, a study was undertaken to establish desorption isotherms for the husks, shells, and kernels of hybrid hazelnuts grown in the Upper Midwest. Clusters were hand-picked from shrubs in Wisconsin and immediately placed in 18 different controlled environments (six different relative humidity levels at three different temperatures). Actual moisture conditioning took place over saturated salt solutions in specially fabricated biomaterial moisture conditioning units. After a six-week period during which the clusters reached equilibrium with their environment via desorption, they were separated into husk, shell, and kernel fractions and returned to their respective conditioning units. After another six weeks in the conditioning units, the moisture content (MC) of each fraction was determined by oven-drying at 103°C for 48 h. Under equilibrium conditions, the kernel MC was found to be only 37% of that for shells, whereas the equilibrium moisture content (EMC) values for husks were on average 14% greater than those for shells. On a dry basis, the average cluster mass was 32.9% husk, 43.9% shell, and 23.2% kernel. Likewise, on a dry basis, the average whole nut mass was 65.5% shell and 34.5% kernel. The desorption data were fit to the Modified Henderson, Modified Chung-Pfost, Modified Halsey, Modified Oswin, and Modified GAB equations. Overall, the best fit to the experimental data was provided by the Modified Chung-Pfost equation with parameters determined using equilibrium relative humidity (ERH) as the dependent variable in regression analyses. For ERH values above 0.70, the temperature-modified form of the GAB equation is recommended for predicting desorption EMC values for hazelnut fractions. Keywords: Desorption, Equilibrium moisture content, Equilibrium relative humidity, Hazelnuts, Kernels, Nuts, Shells, Water activity.