In environment studies of arid areas, erosion and sedimentation are of great importance due to specific climatic, vegetation and soil characteristics. This research aimed to present the mathematical relationships between the Sediment Delivery Ratio (SDR) and 30 influencing factors, including climatic, hydrological, physiographic, vegetation cover, soil and geological factors, in small and arid watersheds of the central desert of Iran. Erosion was determined using the empirical Erosion Potential Method (EPM). From dividing the sediment yield, based on available data, by the calculated erosion, the observed SDR was calculated. Stepwise regression was used to identify the relationships between SDR and the influencing factors. To evaluate the efficiency of the models, the Model Efficiency (ME), Root Mean Squared Error (RMSE) and Coefficient of determination (R2) were calculated. The results showed that the specific erosion and the observed SDR were between 1.67 and 9.55 t/ha-1 yr-1 and the observed SDRs were between 0.08 and 0.88 respectively. Based on ME, RMSE and R2 values of 0.96, 0.18 and 0.95 respectively, a multivariate nonlinear model with variables including slope, area, silt and rainfall intensity at the time of concentration was the most suitable model for estimating SDR. Furthermore, to determine the accuracy and validation of the models, SDR was estimated in two control watersheds using the designed models. The relative error percentages of the estimated SDR by the multivariate nonlinear model were 38.51 and 23.5. Therefore, this model can be used to estimate SDR in other small and arid watersheds of central Iran and similar areas.