Sample treatment procedure, and the experimental parameters that need to be considered in size analysis of diesel soot particles using field-flow fractionation (FFF), are discussed. A probe-type sonication of 10 min in water containing 0.05% of Triton X-100 provides well-dispersed suspension of the soot particles. Sample treatment is found to be important for reduction in particle-particle interaction and, thus, for preparation of well-dispersed soot suspension. FFF elution profile of untreated suspension has a long tailing that does not return to the baseline until the external field is removed. The scanning electron micrographs show that the untreated suspension contains aggregated particles. Extraction of chemicals is also found to be important for preparation of well-dispersed suspension, The suspensions prepared without the extraction step, have broader and larger size distributions than those prepared with the extraction step, due to the presence of soluble organic fractions (SOF) that cause particle-aggregation. FFF provides an excellent repeatability in size analysis of soot particles. The standard deviation of the mean diameter measured by FIFFF was 3.6%. Both Flow FFF (FIFFF) and sedimentation FFF (SdFFF) are tested for size analysis of soot particles. SdFFF provides higher size-selectivity than FIFFF and is less problematic in terms of the particle-channel interaction. Unlike in SdFFF, FIFFF analysis does not require the density information and, thus, the fractogram can be directly transformed to the size distribution. But FIFFF provides lower size-selectivity and has narrower dynamic range than SdFFF.