AbstractFive methods were modified to test for interactions between binder and filler of composite solid propellants. Methods based on uncured binder or model compounds were rate of solution of binder from composite mixtures; centrifugal separation of binder from the filler; of composite mixtures; measurement of the contact angle between binder and filler; and adsorption of binder or model compounds by filler from solution. Stress–strain–birefringence was measured on cured binders containing small amounts of filler. In addition to the polymers used as binders and the model compounds, three physical forms of aluminum, and the additives tris [1‐methylaziridinyl] phosphine oxide (MAPO) and a polysebacate of methyl‐N‐diethanolamine were included in the study. The filler was ammonium perchlorate in all experiments. The polysebacate was an effective adhesion improver in the polyurethane–ammonium perchlorate composite. MAPO was not as effective in the system polybutadiene–ammonium perchlorate. Fibrous forms of aluminum result in a weaker propellant then does powdered aluminum. Nonfunctionally terminated polymers were poorer in adhesion, contact angel, and adsorption tests compared with carboxyl‐ and hydroxyl‐terminated types. The nonpolymeric model compounds yielded