The purpose of this study was to predict the ratio of immunogold labeling of deplasticized epoxy sections and LR-White sections on the basis of theoretical considerations. Tissues used in the experiments were pancreas, pituitary, kidney, thyroid, and fibrin. Antigens used as test proteins were glucagon, somatostatin, thyroglobulin, chromogranin A, ACTH (= Adrenocorticotropt hormone), amyloid A, and fibrinogen. These are proteins of different sizes. The quotient labeling(deplasticized)/labeling(LR-white) was deduced theoretically and compared to measurements based on immunogold experiments to obtain a theoretical model with acceptable correlation to the measurements. This study describes a theory-expressed mathematically-for what happens at the molecular level in immunoelectron microscopy at the surface of deplasticized epoxy sections and acrylic sections. The theory explains why we normally get about the same amount of immunogold labeling using LR-White sections (acrylic resin) and deplasticized epoxy sections. Taking the nuances into account, the theory indicates increased usefulness of deplasticized epoxy sections when the diameter of the protein carrying the epitopes decreases. Copyright (C) 1996 Elsevier Science Ltd.