Good soil structure is important for crop growth. One of the main characteristics of the soil structure, which affects the plant growth and its yield, is soil aggregate size distribution. There are several ways to characterize the stability of soil aggregates such as the mean weight diameter (MWD). Researchers generally use MWD as the most important criterion to express the degree of soil fragmentation of a loamy sand soil. In this study, the effects of soil moisture content at two levels, tractor forward speed at five levels and working depth at five levels were investigated on soil fragmentation for seedbed preparation using a combination of primary and secondary tillage implements of subsoiler, moldboard plow and disk harrow as conventional tillage combination in the region. Field experiments were carried out in the research field of University of Mohaghegh Ardabili based on a completely randomized design (CRD) with three replications. In this paper, an intelligent model based on Mamdani fuzzy model was developed to predict soil fragmentation during tillage practice. The model included 50 laws. Analysis of variance showed that the main effects of travel speed and working depth were significant on the soil fragmentation (at probability level of 1%). By increasing the forward speed, working depth and soil moisture content, MWD of aggregates significantly increased. Moreover, the interactive effect of speed and working depth was significant (at probability level of 1%). Also, using the fuzzy model, the optimum range of the investigated parameters was achieved 12-16% for moisture content of a loamy sand soil, 5.92 km/h for travel speed and 15 cm for working depth.