The Free Tropospheric Experiment (FREETEX’98) was conducted at the Jungfraujoch Observatory in the Swiss Alps (3580 in above sea level) during the well‐documented spring maximum in ozone. In spring the Jungfraujoch frequently lies in the free troposphere but can also be influenced by air from the planetary boundary layer. Measurements of NOx, NOy, peroxy‐acetylnitrate (PAN), HCHO, O3, CO, nonmethane hydrocarbons, peroxy radicals, j(O1D), j(NO2), and a variety of other tropospheric constituents crucial to ozone photochemical cycles were made over a 1‐month period. Two independent measurements of NOx, NOy, and PAN showed good agreement. Average free tropospheric daytime NO levels were about 50 pptv, sufficient to sustain photochemical ozone formation. Although high mixing ratios were encountered, PAN decomposition did not contribute to NOx production during FREETEX’98. Ozone production efficiencies (E N) derived from observed ΔO3/(NOz) ratios in free tropospheric air were 20–30 molecules of O3 produced per NOx molecule oxidized and agreed well with a photochemical model. A much lower ozone production efficiency of 4 was determined in a photochemically aged air mass arriving from southern Europe, in line with other measurements and calculations in regimes containing high levels of oxidized nitrogen. Model simulations indicated that by sequestering NOx and HO2, low‐temperature formation of peroxynitric acid (PNA) decreased ozone production by 20% and instantaneous ozone production efficiencies by 40%, whereas PAN formation had little effect. The model reproduced well the observed sharp transformation from ozone production to ozone destruction (defined as ΔO3/Delta;(NOz) = 0) at 20–25 pptv NO. The observed and calculated strong dependence of EN on NOx concentration in the low‐NOx regime highlights the difficulty in assigning a single O3 production efficiency value to remote regions, where most of the CO and CH4 in the atmosphere are oxidized.