Measurements of NOx (NO+NO2) and NOy (total reactive nitrogen oxides) during February--April 1996 at the northern tip of Newfoundland are used to determine levels in the local marine boundary layer (MBL) and assess the adequacy of current understanding of the processes controlling NOx levels over the northern North Atlantic, as expressed through previously reported simulations using the Geophysical Fluid Dynamics Laboratory (GFDL) Global Chemical Transport Model (GCTM). Median mixing ratios of NOx and NOy in the local MBL were 24 parts per trillion by volume (pptv) and 200 pptv, respectively. These levels are 35--64% above background levels measured in the remote MBL in summer and fall during the Chemical Instrumentation Test and Evaluation (CITE 2), North Atlantic Regional Experiment (NARE-93), and Pacific Exploratory Mission-West (PEM-West) A measurement campaigns and are similar to or somewhat higher than anthropogenically influenced levels observed in winter-spring during the PEM-West B campaign. The magnitude of median NOx and NOy levels in the local MBL is not due to events with high reactive nitrogen oxides levels. Instead, these relatively high median levels are likely the result of dispersion of anthropogenic emissions over a large region. A detailed comparison with results from the GFDL GCTM indicates that measured March and April average NOx levels are significantly lower than simulated levels over the north central North Atlantic. The frequency and magnitude of modeled and observed elevated-NOx events were similar, indicating that the conditions responsible for relatively direct long-range transport events were similar. This indicates that interannual variability probably did not cause the discrepancy in monthly average NOx values. However, simulated elevated NOx events are much longer than are observed. This difference appears to be at least partially responsible for the higher average NOx values simulated by the model. These results indicate that model-based estimates of this region's contributions to the global ozone budget may be too high. Accuracy of the NOx measurements is estimated to be 6%, while conservative analysis of conversion efficiencies indicates a negative bias of ≲18% in the determination of gas-phase NOy compounds. ¿ 1999 American Geophysical Union