UARS observations bearing upon the NOx/NOy and NOx/HNO3 ratios were revisited within the framework of advances in laboratory studies that were spurred by the findings, around 1999, of considerable model underestimation of the observations. The objective was to study the two models of the OH, NO2 association reaction, one that considers the single HNO3 product channel and the other that considers dual HNO3 and HOONO products channels. Rate coefficients of the OH + HNO3 and the single-channeled OH + NO2 → HNO3 reactions recommended in JPL00-3 produce remarkably improved agreement between the modeled and observed NOx/HNO3 and NOx/NOy, compared to the case with the recommendations in JPL-97. This agrees with the results of other previous studies, such as those using the Photochemistry of Ozone Loss in the Arctic Region in Summer and MkIV data. The consumption of OH, NO2 pair in the lower stratosphere via the three-body association occurs at almost the same rate (within ¿5%) whether the single- or the double-channeled version of that process is used with rate coefficients from JPL00-3. However, only 70--80% of this consumption yields HNO3 in the dual-channeled version, compared to 100% in the case of the single-channeled version. Consequently, the dual-channeled model of HNO3 formation performs better than the single-channeled model in predicting NOx/HNO3 and NOx/NOy ratios. At 46.4 mbar, for example, the mean percentage difference between the modeled and the observed NOx/HNO3 equals about -10%, if HOONO is included. The corresponding number is about -19% when HOONO is excluded. Because laboratory experiments on OH, NO2 association that directly detect HOONO isomer are still maturing, further studies are needed. Further studies are also needed to understand the intriguing tendency of the model-measurement differences to maximize around 20 mbar.