Aerosol behavior is a central topic in the overall determination of source terms of accidents in nuclear power plants. Aerosol behavior codes have been developed for ∼20 yr and their state of development is considered to be well advanced. The acceptance of results of the calculations, however, was poor as far as physical understanding and the results of basic research were involved. Proof that the codes are complete and the choice of models is adequate for a situation during an accident required large-scale integral experiments that simulated real conditions as closely as necessary and possible. Recently, three large-scale experimental programs were carried out dealing with the behavior of aerosols during core-melt accidents in light water reactors (LWRs). In the Nuclear Safety Pilot Plant (NSPP) program, the principal behaviors of different insoluble aerosols and of mixed aerosols were measured in dry air atmospheres and in condensing steam-air atmospheres contained in a 38-m3 steel vessel. The Demonstration of Nuclear Aerosol Behavior (DEMONA) program used a 640-m3 concrete containment model to simulate typical accident sequence conditions, and measured the behavior of different insoluble aerosols and mixed aerosols in condensing and transient atmospheric conditions. Part of the LWR Aerosol Containment Experiments (LACE) program was also devoted to aerosol behavior in containment; an 852-m3 steel vessel was used, and the aerosols were composed of mixtures of insoluble and soluble species. The results of these experiments provide a suitable data base for validation of aerosol behavior codes. Fundamental insight into details of aerosol behavior in condensing environments has been gained through the results of the NSPP tests. Code comparisons have been and are being performed in the DEMON A and LACE experiments.