Soil texture is agronomically important because it affects the amounts of nutrients, organic matter, water and air in the soil. It is also important in modelling the behaviour of soil in terms of erosion, solute transport, pesticide adsorption etc. Soil surveyors traditionally assess texture by hand-texturing in the field and classify it based on the textural triangle, not the exact proportions of sand, silt and clay present. The pipette method has been used to assess soil texture in the laboratory for many years. It is based on Stoke's law of streamlined flow and relies on the fact that spherical particles of a certain size take different amounts of time to settle. Finally, laser methods are based on the assumption that a laser directed at a soil sample is refracted by different amounts by different sized spherical particles. The degree of laser refraction is registered at photoreceptors for various sized particles. The three methods of particle size determination described usually give similar results when properly implemented, but there are some additional discrepancies that arise when determining the texture of soil formed on chalk. A survey of several field sites in the U.K. showed discrepancies between soil textures obtained from maps of previous soil surveys and those determined by laser methods. The soil for one field was developed on the Lower Greensand. The analysis of selected samples and reference soil samples by the pipette method showed that the soil surveyor had over-estimated the contrast in texture between two mapping units by hand-texturing. At the second site the soil had developed on the Upper Chalk and associated clay with flints deposits, and the textural classes by laser diffraction were different from those of a previous survey. There were also improbable relationships between the different particle size fractions and other soil properties. For example, the percentage of organic matter and volumetric water content were negatively related to percentage clay and positively related to percentage sand. These improbable relationships were most pronounced where the soil was most shallow on a steep slope and was formed directly on the chalk rather than in the surficial deposits. Selected samples were re-analyzed for particle size using the pipette method and the percentage of CaCO3 was also determined using the LECO carbon method. These results confirmed that the discrepancies between the different methods of particle size determination were greatest where the soil was most calcareous. The Upper Chalk is very pure and is made up almost entirely of tiny calcite crystals. These can dominate the clay-sized fraction, yet they do not have the same characteristics in terms of water retention as typical phyllo-silicate clays minerals. Soil containing calcium carbonate is known to feel more silty when hand-texturing than it might actually be. The pipette method usually uses an average value for the particle density in determining settling times for particles of a given size. Perhaps this should be changed when the mineral fraction is almost entirely made up of calcite crystals. Finally, the pre-treatments involved in dispersing the soil for the pipette method and particularly laser methods can result in the break-up of chalk fragments into smaller calcite crystals making the soil appear more clayey in texture than it is. This research has shown that there are clear problems with the commonly used methods when the soil is highly calcareous. Carbonate material could be removed before particle size determination, but if this is a large proportion of the mineral fraction as for chalk soil, this is unsatisfactory. A new approach to particle size determination on chalk soil is needed. As an interim measure for agronomic studies, hand-texturing seems to give the best indication of how the soil will behave under agricultural management even if is does not give the exact proportions of each particle size class.