The paper presents a summary of the results of cyclical triaxial load testing of samples of silt and silty sands. The paper emphasizes differences in behaviour observed between reconstituted triaxial samples of clean sand, of sand containing 10, 20, 30 and 60% silt, pure silt, and undisturbed samples of silt and silty sand. An important observation is that the mechanisms of deformation for sit are different for reconstituted and undisturbed samples, the undisturbed sample having a specific ‘geological’ structure which seems to slow down the excess pore water pressure accumulation, but which still results in cyclic deformations regularly increasing from the very beginning of the test and rapidly reaching high levels. The other important observation is that fine-grained noncohesive soils such as silts and silty sands can be as, or even more, susceptible to liquefaction as clean sands. Test results on samples of sand containing 10, 20 or 30% of silt indicate lesser resistance to liquefaction than pure sand samples. The paper shows the difficulty in identifying a representative parameter to compare the behaviour of silts and silty sands with pure sand, and it seems that more research will be needed in this area.