As an important environmental index, grain-size character of lake sediments has been widely used in the lake sediment and environmental change studies, accurate measurement is an indispensable requisition for its application.In the depositing process of the lake sediments, chemical and physical change can be taken place and some organic matter, some carbonate may form, organic matter and carbonate can make the sediments cement, and the cementation is not helpful for the grain-size measurement, so pre-treatments is very necessary to get accurate grain-size data.There are some differences in the pre-treatments of lake sediments measurement in differeut laboratories.Here we present our experimental data of lake sediment grain-size measurement from two different lakes.Based on different pre-treatments, the results of the measurements were anaJyzed and compared, and two preferable pre-treatment methods were presented.The results indicate that the sediments from different lakes have different responds to different pre-treatment, it can be seen from the whole samples that the pre-treatment that dipping in the distilled water and dispersed by ultrasonic(method F-10 in the text) obtained the best dispersing effect, and the traditional pre-treatment (method E-3 in the text), which is also the most common use one, also got rather good results.Whereas pre-treated by (NaP₃)₆ could make the sediment particles easily to be shattered by ultrasonic and then another very thin particle peak appeared, possibly (NaP₃)₆, as a kind of dispersant, decreasing the flocculation effect between tlie particles and then the direct striking effect to the particles of the ultrasonic strengthened, under the strong striking, particles can be broken up.We believe that the bimodal distribution curves do not reflect the real condition of the samples, and this kind of abnormal distribution can bring false environmental information, more attention need to pay to this phenomena when measuring the grain-size of the lake sediments.