As farmers in southern Australia typically apply nitrogen (N) to cereal crops by top-dressing with ammonia (NH 3 ) based fertilizer in late winter or early spring there is the potential for large losses of NH 3 . This paper describes the results of micrometeorological measurements to determine NH 3 loss and emission factors following applications of urea, urea ammonium nitrate (UAN), and ammonium sulfate (AS) at different rates to cereal crops at two locations in southern Australia. The amounts of NH 3 lost are required for farm economics and management, whilst emission factors are needed for inventory purposes. Ammonia loss varied with fertilizer type (urea [ UA-N [ AS) and location, and ranged from 1.8 to 23 % of N applied. This compares with the emission factor of 10 % of applied N advocated by IPCC ( 2007). The variation with location seemed to be due to a combination of factors including soil texture, soil moisture content when fertilizer was applied and rainfall after fertilizer application. Two experiments at one location, 1 week apart, demonstrated how small, temporal differences in weather conditions and initial soil water content affected the magnitude of NH 3 loss. The results of these experiments underline the difficulties farmers face in timing fertilization as the potential for loss, depending on rainfall, can be large.