Observations of OH and HO2 from Aura MLS for four seasons and diurnal profiles from the FIRS-2 balloon instrument for Fall 2004 are compared with photochemical model simulations testing three sets of kinetics parameters. MLS and FIRS-2 OH profiles, between 25--60 km, are lower than model results using standard kinetics. Use of a faster, previously published rate constant for O+OH leads to better agreement with MLS and FIRS-2 profiles of OH. A 20% increase in the rate of HO2+OH and the faster rate for O+OH results in improved overall agreement with observations of OH, HO2, HOx, and HO2/OH. Since the MLS and FIRS-2 observations of HOx are reasonably well described by these models, they are therefore not consistent with the previously reported HOx dilemma. However, all models considered here result in calculated odd oxygen loss exceeding production, consistent with the long standing ozone deficit problem.