In this study, the effects of drying on Atterberg limits is investigated using 30 soil samples taken from 0.2 m to 4 m depths, from nine different landslide sites in pyroclastic deposits in Northern Turkey. Samples for Atterberg limits are prepared by (i) starting from in-situ moisture content without any drying, (ii) drying at 60 degrees C and (iii) drying at 110 degrees C. The range of liquid limit values is 35%-103% and plasticity index is 0%-37% when samples are prepared from their in-situ moisture content. Liquid limits of the samples, when they are prepared by drying at 110 degrees C oven, are 1%-30% lower than the liquid limits prepared from moist condition. The mineralogy of the samples is investigated via X-ray diffraction analyses and Scanning Electron Microscopy photographs, indicating existence of halloysite mineral. Organic contents of the samples, determined by loss on ignition test, is in the range of 4.3%-12.1%. It is concluded that drying significantly influences the results of Atterberg limits if the sample contains some organic material and/or halloysite mineral. Both liquid limit and plasticity index values of the soils decreased as the sample preparation method changed from moist preparation to drying at 110 degrees C. Existing soil classification methods may fail to represent the influence of organic matter on soil behaviour for soils containing < 15% organic material.