Results from an airborne intercomparison of techniques to measure tropospheric levels of nitrogen dioxide (NO2) are discussed. The intercomparison was part of the National Aeronautics and Space Administration's Global Tropospheric Experiment and was conducted during the summer of 1986. Instruments intercompared included a two-photon nitric oxide (NO) laser-induced fluorescence system with laser photolysis of NO2 to NO, and NO/O3 chemiluminescence detector using FeSO4 for conversion of NO2 to NO, and NO/O3 chemiluminescence detector with arc lamp photolysis of NO2 to NO, and a tunable diode laser multipath absorption system. All intercomparisons were for NO2 mixing ratios of <200 pptv with most at mixing ratios of <100 pptv. The FeSO4 converter was found to convert peroxyacetyl nitrate (PAN) to NO, resulting in NO2 values a factor of 2--3 higher than reported by the other techniques. Thus the FeSO4 converter data are excluded from the analyses. Intercomparison data were analyzed in various mixing ratio ranges. Good correlation was observed between the remaining three instruments for those data sets which included mixing ratios to 100 or 200 pptv, showing on the average a 30--40% level of agreement among the techniques. However, when the data were restricted to mixing ratios of <50 pptv, little correlation among the measurements was observed. Even though correlations were poor at mixing ratios of <50 pptv, the tunable diode laser system tended to be high compared to data reported by the two-photon laser and arc lamp chemiluminescence systems, and agreement between the latter two instruments was generally better than 20 pptv with an equal tendency for one to be high relative to the other. ¿ American Geophysical Union 1990