This paper describes an experimental study of the effects of sampling disturbance in an Australian natural soft clay and the consequences of different sample quality on the representativeness of soil parameters used in geotechnical designs. The paper is divided into three sections. Laboratory test results obtained from specimens retrieved using three different tube samplers as well as the Sherbrooke (block) sampler are first described. Then, the sample quality assessment, using available indices proposed for soft soils, is presented. It is shown that sample quality varies with the stress paths and boundary conditions applied in laboratory tests. Finally, mechanical soil properties derived from specimens retrieved using the different samplers are used in the prediction of two classical problems in soil mechanics: the settlement and excess pore pressure response underneath an embankment as well as the settlement and bearing capacity of a shallow footing. These two examples are used here to highlight the consequences of poor sampling in practice.