Ammonia (NH 3 ) emissions from gasoline-fueled vehicles have become an important source of pollution affecting urban air chemistry. NH 3 influences the acidity of atmospheric depositions and it is involved in secondary aerosol formation. NH 3 has to be considered as a secondary pollutant of the three-way-catalyst (TWC), since it is formed de novo during the DeNOx process. The extent of traffic-related hydrogen (H 2 ) emissions and its impact on atmospheric redox chemistry is not well understood but is of increasing importance when we develop towards a hydrogen-based society. Herein we report on tail-pipe H 2 , NH 3 , and NO emissions of gasoline-fueled Euro-3 passenger cars at transient driving from 0 to 150kmh −1 . The effects of velocity, acceleration, deceleration, and cold start were deduced from time-resolved EI- and CI-MS data. On a molar basis, H 2 emissions were always higher than those of NH 3 and NO by about an order of magnitude. H 2 and NH 3 emissions are correlated to some degree, as soon as catalyst light-off occurred. NH 3 emissions exceeded those of NO for most vehicle conditions. Mean NH 3 /NO mixing ratios around two were observed with the exception of the cold start, where NO was present in large excess. Catalyst light-off is indicated by a fast transition from a NO- to a NH 3 -rich exhaust gas. All emissions clearly depend on speed and acceleration. Mean velocity-dependent emission factors varied by about one order of magnitude from 17 to 720, 8 to 170, and 7 to 80mgkm −1 for H 2 , NH 3 , and NO, respectively, with emission minima for all three pollutants when driving 70–90kmh −1 . We conclude that the investigated Euro-3 vehicles are mainly operated under slightly reducing conditions, where NH 3 and H 2 emissions dominate over those of NO. Under these conditions, all vehicles fulfill the valid emission limit for NO x .