[en] Quantification of radioactive contamination in the environment is often desirable so that health implications can be determined and appropriate remedial measures taken. In-situ gamma spectrometry is a potentially powerful technique that has the advantage of speed and spatial averaging over a large area. Calibration, however, is complex not least because the distribution of the activity in the field is generally unknown. To overcome this problem, methods have previously been proposed to derive information about the vertical activity distribution. All of these methods assume a fixed simple activity distribution model, the parameters of which are derived from varying in-situ spectral region ratios. In this work, these methods have been extensively assessed and compared both theoretically and experimentally for the quantification of ¹³⁷Cs contamination. In all cases, the best method was that which used a lead plate to alter the contributing angular interval. Experimentally, this Lead Plate Method predicted the activity concentration to within a factor of between 1.50 and 1.66 on average. A new method, using collimated detector measurements within a shallow well was developed and theoretically optimised in terms of number and depths of detector positions and type of collimation. This optimised Submerged Detector Method was assessed and compared with the Lead Plate Method both theoretically and experimentally. The latter comparison involved 19 sites across Wales, UK where there is a wide range of ¹³⁷Cs activity levels and vertical distributions. The Submerged Detector Method was found to be more accurate both in the theoretical modelling assessment and in the experimental comparison, where the method was found to predict the activity concentration to within a factor of 1.35 on average. Distribution profiles predicted by the Submerged Detector Method were also found to be generally closer to the true profiles than those obtained using the Lead Plate Method. (author)