Global chemistry transport models often overestimate the HNO3/NOx ratio in the free troposphere compared with observations. The sensitivity of the HNO3/NOx ratio to the parameterization of washout is studied using two global models, IMAGES and MOZART. The observed ratio of HNO3/NOx is about 4¿3 in the free troposphere over remote areas. Calculations with the IMAGES model are in good agreement with these observations, but those with the MOZART model are too high (about 7--20). The good agreement found for IMAGES might be fortuitous, caused by the physically unrealistic parameterization of washout as a continuous removal rate. Applying different magnitudes of a continuous washout or different intermittent washout frequencies increases the HNO3/NOx ratio by about a factor of 2--3, worsening the agreement with observations. Smaller, but significant, changes of the order of 10--20% can be seen in NO, NO2, and ozone as a result of changes in the washout parameterization. The heterogeneous conversion of HNO3 on SO4 (used as a proxy for soot) yields only a small improvement in the HNO3/NOx ratio for most remote locations. Significant effects of this heterogeneous conversion can be expected over and downwind of polluted locations. Given the sensitivity of the HNO3/NOx ratio to the parameterization of washout, this ratio may be more suited to test the quality of the parameterization of the washout than the nitrogen chemistry in global chemistry transport models. Expressed differently, the HNO3/NOx ratio is little affected by and not very important for tropospheric chemistry in these models. ¿ 2001 American Geophysical Union