A reexamination of data and theoretical computations for the title reaction leads to the conclusion that both HONO2 (nitric acid) and HOONO (pernitrous acid) can be formed. We describe hindered-Gorin RRKM calculations that fit most of the extant data and explain differences among the studies. We conclude that the rate constant for nitric acid formation at low temperatures could be considerably (similar to 20%) lower than the data for loss of OH in the presence of NO2 would indicate. We have examined the atmospheric consequences of this conclusion on ozone and nitrogen oxide concentrations through the use of box models. We include a tentative recommendation for the rate constants for the individual steps as functions of temperature and density in the NASA format.