A pulsed stand-off Raman system has been built and optimised for the qualitative and quantitative analysis of inorganic and organic samples including explosives. The system consists of a frequency doubled Q-switched Nd:YAG laser (532 nm, 10 Hz, 4.4 ns pulse length), aligned coaxially with a 6″ Schmidt–Cassegrain telescope for the collection of Raman scattered light. The telescope was coupled via a fibre optic bundle to an Acton standard series SP-2750 spectrograph with a PI-MAX 1024RB intensified CCD camera equipped with a 500-ps gating option for detection. Gating proved to be essential for achieving high signal-to-noise ratios in the recorded stand-off Raman spectra. In some cases, gating also allowed suppression of disturbing fluorescence signals. For the first time, quantitative analysis of stand-off Raman spectra was performed using both univariate and multivariate methods of data analysis. To correct for possible variation in instrumental parameters, the nitrogen band of ambient air was used as an internal standard. For the univariate method, stand-off Raman spectra obtained at a distance of 9 m on sodium chloride pellets containing varying amounts of ammonium nitrate (0–100%) were used. For the multivariate quantification of ternary xylene mixtures (0–100%), stand-off spectra at a distance of 5 m were used. The univariate calibration of ammonium nitrate yielded R2 values of 0.992, and the multivariate quantitative analysis yielded root mean square errors of prediction of 2.26%, 1.97% and 1.07% for o-, m- and p-xylene, respectively. Stand-off Raman spectra obtained at a distance of 10 m yielded a detection limit of 174 μg for NaClO3. Furthermore, to assess the applicability of stand-off Raman spectroscopy for explosives detection in “real-world” scenarios, their detection on different background materials (nylon, polyethylene and part of a car body) and in the presence of interferents (motor oil, fuel oil and soap) at a distance of 20 m was also investigated.