Data illustrating the performance characteristics of a proton transfer reaction-mass spectrometer (PTR-MS) under both laboratory and field conditions are presented. Under laboratory conditions, we demonstrate that PTR-MS measures (within 10%) a 2.6 ppbv concentration of gaseous dimethyl sulfide. Using a stepwise dilution of a gaseous isoprene standard, we demonstrate the linearity of the response of PTR-MS across 3 orders of magnitude of mixing ratios, from 100 ppbv to less than 100 pptv. By combining this data set with that of its monosubstituted 13C isotopic analogue, we demonstrate the ability of the instrument to reliably measure concentrations as low as 50 pptv and to detect concentrations at significantly lower levels. We conclude our laboratory characterization by investigating the components of the instrument noise signal (drift, mean, and range) and develop an expression (noise statistic) that reliably predicts the instrumental noise associated with any signal across a wide range of masses. In the field, we deployed a PTR-MS at a clean-air coastal site and an urban kerbside monitoring station to demonstrate the measurement of atmospheric dimethyl sulfide and benzene concentrations, respectively. At both sites, we were able to monitor diurnal variations in concentrations at unprecedented temporal resolutions (<5 min between successive measurements). We then demonstrate how the noise statistic can be applied to enable real fluctuations in atmospheric VOC concentrations to be reliably distinguished from instrument noise. We conclude by demonstrating how PTR-MS can be used to measure real-time VOC emission rate changes from vegetation in response to external forcing by examining the effect varying photon-flux density has upon emissions of isoprene from a Sitka spruce tree.