We have studied the NH2 + NO reaction theoretically in order to try to deduce a theoretical "model" that will accurately reproduce both the total rate coefficient k(T)(T) and the branching fraction alpha(T) of the reaction NH2 + NO --> N-2 + H2O (a), NH2 + NO --> NNH + OH (b), and NH2 + NO --> N2O + H-2 (c), where k(T) = k(a) + k(b) + k(c) and alpha = k(b)/k(T). The analysis, which makes the RRKM assumption and utilizes conventional transition-state theory for the internal-rearrangement transition stales and microcanconical/fixed-J variational transition-state theory for the bond fissions, is discussed at length. The results of the analysis show clearly that k(T)(T) is determined almost exclusively by the transition state for the 1,3 hydrogen transfer connecting the initial NH2NO complex to HNNOH. The branching fraction is sensitive to several features of the potential energy surface, most of them associated with the fragmentation of the various HNNOH complexes into NNH + OH. By adjusting properties of the potential energy surface, we have constructed a theoretical model that predicts results for both k(T)(T) and alpha(T) that are in good agreement with experiment. A variety of sensitivity analyses for the branching fraction indicate that reaction b is most likely thermoneutral to within +/-1 kcal/ mel. Our prediction of k(c)(T) is between 2 and 3 orders of magnitude smaller than values deduced from experiment, suggesting that the experiment may have detected the existence of a fourth channel, HNNO + H, or may have been contaminated by secondary reactions.