A new and holistic modelling methodology is developed to investigate the impact of line-edge roughness (LER) on time-dependent dielectric breakdown (TDDB). Unlike the conventional approach with a normal distribution for LER profiles [1-3], a realistic simulation of line-edge roughness (LER) profiles is carried out based on a power-spectrum density (PSD) function including the effects of both roughness and correlation-length (σ and λ). In contrast to the previous work, we show that LER roughness causes a significant reduction not only in T63 (scale parameter) but also in β (shape parameter). While our results show that the thickness dependence of T63 and β with LER follows a similar trend of the conventional percolation model, unlike porosity effect introduced in conventional percolation model the effective defect size (a0) is found to increase with increasing roughness (σ). To examine the strong impact of larger roughness on TDDB, a field-enhancement effect is included in LER modelling. Although the impact of field-enhancement due to protrusions on TDDB has been qualitatively discussed in the past, our LER simulation with a statistical approach has achieved quantitative agreement with experimental data, revealing an in-depth knowledge for the separate roles of LER profiles and field-enhancement. Thus, our methodology can be implemented as a realistic reliability assessment tool for technology qualification.