The autoxidation of NO was studied in glass-like matrices of 2-methylbutane at 110 K and in a 8:3 v/v mixture of 2,2-dimethylbutane and n-pentane (rigisolve) at 80-90 K, by letting gaseous NO diffuse into these solvents that were saturated with O. In 2-methyllbutane, we observed a red compound. However, in rigisolve at 85-90 K, a bright yellow color appears that turns red when the sample is warmed by 10-20 K. The new yellow compound is a precursor of the red one and also diamagnetic. The UV-vis spectrum of the yellow compound contains a band which resembles that present in ONOO. Because the red and yellow intermediates are not paramagnetic, we postulate that O═N-O-O is in close contact with NO, or with another O═N-O-O. Diffusion of gaseous O into rigisolve saturated with NO does not produce a color; however, a weak EPR signal (g = 2.010) is observed. This signal most likely indicates the presence of ONOO. These findings complement our earlier observation of a red color at low temperatures and the presence of ONOO in the gas phase (Galliker, B.; Kissner, R.; Nauser, T.; Koppenol, W. H. Chem. Eur. J. 2009, 15, 6161-6168), and they indicate that the termolecular autoxidation of nitrogen monoxide proceeds via the intermediate ONOO and not via NO.