A large set of isoprene and isoprene oxidation product concentration data from four North American sites was examined to assess the NOx dependence of the daytime oxidation of isoprene. Sites that represent a wide range of NOx (50 ppt to 30 ppb) were studied and include the Dickson, Tennessee, and Cornelia Fort Air Park sites during the 1999 Southern Oxidants Study, the Pellston, Michigan, site during the 1998 PROPHET summer intensive, and the Kejimkujik National Park site during the Atlantic 1996 study. Knowledge of NOx and HOx concentrations were critical for this evaluation. While NOx data are readily available at all sites, HOx data are limited. We employed a simple 10-reaction HOx model to calculate steady state OH radical concentrations as a function of [NOx] to enable analysis of the data from all sites. Here, we use methyl vinyl ketone (MVK) concentrations to quantify the extent of isoprene-OH oxidation. Making use of the MVK/isoprene ratio, we show that the rate of production of isoprene oxidation products at various North American sites, although highly variable, exhibits the crossover from NOx-dependent to VOC-dependent conditions at 8 ppb [NOx], in agreement with what is calculated from HOx measurements and our calculations.