OH and HO2 were measured with the Airborne Tropospheric Hydrogen Oxides Sensor (ATHOS) as part of a large measurement suite from the NASA DC-8 aircraft during the Intercontinental Chemical Transport Experiment-A (INTEX-A). This mission, which was conducted mainly over North America and the western Atlantic Ocean in summer 2004, was an excellent test of atmospheric oxidation chemistry. The HOx results from INTEX-A are compared to those from previous campaigns and to results for other related measurements from INTEX-A. Throughout the troposphere, observed OH was generally 0.95 of modeled OH; below 8 km, observed HO2 was generally 1.20 of modeled HO2. This observed-to-modeled comparison is similar to that for TRACE-P, another midlatitude study for which the median observed-to-modeled ratio was 1.08 for OH and 1.34 for HO2, and to that for PEM-TB, a tropical study for which the median observed-tomodeled ratio was 1.17 for OH and 0.97 for HO2. HO2 behavior above 8 km was markedly different. The observed-to-modeled HO2 ratio increased from ~1.2 at 8 km to ~3 at 11 km with the observed-to-modeled ratio correlating with NO. Above 8 km, the observed-to-modeled HO2 and observed NO were both considerably greater than observations from previous campaigns. In addition, the observed-to-modeled HO2/OH, which is sensitive to cycling reactions between OH and HO2, increased from ~1.5 at 8 km to almost 3.5 at 11 km. These discrepancies suggest a large unknown HOx source and additional reactants that cycle HOx from OH to HO2. In the continental planetary boundary layer, the observed-to-modeled OH ratio increased from 1 when isoprene was less than 0.1 ppbv to over 4 when isoprene was greater than 2 ppbv, suggesting that forests throughout the United States are emitting unknown HOx sources. Progress in resolving these discrepancies requires a focused research activity devoted to further examination of possible unknown OH sinks and HOx sources.