Vacuum-ultraviolet/photofragmentation-laser-induced fluorescence (VUV/PF-LIF) has been demonstrated to be a highly specific and sensitive method for the quantitative measurement of atmospheric ammonia (NH3). The fluorescence detected in this approach results from the two photon (193 nm) photofragmentation of NH3 followed by the LIF excitation of the NH(b1(summation)+) yields NH(c1II) (at 452 nm) and the monitoring of fluorescence from the NH(c1II) yields NH(a1(Delta) ) transition at 325 nm. Limits of detection for the instrument presented here are < 10 pptv and < 4 pptv for one and five minute integration periods, respectively, under ambient sampling conditions. The technique is free from interferences and system performance does not significantly degrade under adverse sampling conditions (i.e. rain, fog, clouds, haze, etc.). Spectroscopic selectivity in the NH(b1(summation)+) yields NH(c1II) transition is sufficient to resolve 15NH3 and 14NH3 contributions for use in atmospheric tracer studies. Average ammonia measurements at Stone Mtn., GA, range from approximately equals 110 pptv for air temperatures < 5 degree(s)C to approximately equals 240 pptv for air temperatures >= 5 degree(s)C over the period from December 1987 to the end of April 1988. Ammonia levels measured at Green Mountain Mesa, Boulder CO, ranged from 10 pptv to 10 ppbv for measurements made during March 1989. Ammonia levels were seen to vary from about 300 pptv to greater than 5 ppbv over time scales of < 10 minutes in this latter data set. These results and future instrument improvements are discussed.