Different DC electrical conductivity models have been proposed to predict the properties of composite materials. Among these, the Mamunya model gives a good agreement with experimental data for different carbon black nanocomposites below the threshold concentration. In this paper, we extend the study using the Scarisbrick model. However, the percolation phenomenon is not observed and a modified model which takes into account some parameters, such as the aspect ratio and surface area to volume ratio of fillers, is used to accurately justify the conductive network formation. The modified Scarisbrick model shows clearly the percolation phenomenon and gives satisfactory results for the series samples which present high percolation threshold.