A shape classification of sand particles according to thickness is obtained by sieving a sediment sample through a nest of several screens with slotted openings, the widths of which are smaller in each subjacent screen. The sample is first sieved through a pair of round-holed screens, the openings of which are nearly the same. This produces a subsample narrowly limited in particle size. It is shown that the shape distribution in the sediment fraction obtained through use of the round-holed screens is unbiased. The distributions of particle thickness in the individual slotted screen fractions are shown to be bell-shaped, skewed depending on location relative to the modal class, and overlapping to a degree that is predictable. In using the method, slotted screen fraction weights are converted to number of particles per fraction. These particles are distributed keeping the areas trader the individual distributions proportional to the corresponding number of particles. The numbers of particles per standard class interval of particle thickness are summed over the different distributions to yield the desired distribution of relative particle thickness for the sample.