ABSTRACT With increased interest in fine-grained sediments it is imperative that a firm basis exist for comparative studies of these cohesive sediments. In this study the size distributions of continental slope and rise muds are shown to be dependent on both method and duration of sample pretreatment. Statistical analysis of 171 size distributions indicates that of the four most frequently used sample preparation methods (soak, stir, shake and ultrasonify) the time required to reach a ‘terminal’ distribution beyond which no ‘fine-shift’ was detected varied from 15 minutes to > 90 hours for a given sample solely as a function of preparation method. Data from this study also indicate that sample preparation by simple soaking alone will probably yield fine-grained sediments which, when analysed by pipette or microsedimentation accumulation balance, undergo a continuous change in size distribution during analysis and may never reach a ‘terminal’ distribution during the analysis time. On the other hand stirring and ultrasonification were shown to be the most rapid and consistent methods for obtaining a sample's‘terminal’distribution.