Lightning-related fires and explosions may trigger escalated severe accidents for external floating roof tank (EFRT) farms. Quantitative risk assessment of lightning risk for EFRTs is an effective approach to reduce casualties and property damage. In this research, a quantitative methodology for the risk assessment of direct lightning strike on EFRTs is proposed, in which the risk-attenuating factors are considered by three special sub models. The first sub-model allows the calculations of the frequencies of major accident scenarios with a small amount of inputs. The second sub-model allows the analysis of the influence of uncertainties on the physical effect due to full surface fire. By the third sub-model, the contribution of automatic firefighting systems and evacuation behavior to the decrease of the damage or death probability can be quantified. The three coherent sub-models constitute the main framework of the methodology in this research, and they are also applicable for other quantitative risk analysis. To demonstrate the applicability of this methodology and the flexibility of the sub-models, a case study is investigated and the contributions of risk-attenuating factors are further discussed. Finally, suggestions are proposed to optimize the lightning risk assessment for EFRTs.