This paper investigates the influence of considering actual insulation volt-time curves both in the calculation of lightning-induced voltages and in the estimation of the number of flashovers an overhead wire may experience per year due to nearby lightning strokes. The flashover mechanism is modeled according with the integration method, which is used as a reference for comparisons with the simplified 1.5 CFO flashover criterion traditionally used in the estimation of the lightning performance of overhead distribution lines. Sensitivity analysis show the dependence of flashovers on the shape and front time of the assumed channel-base current. The obtained results suggest that the simplified 1.5 CFO flashover criterion is likely to underestimate the number of flashovers an overhead line may experience per year due to nearby lightning strokes. This result is confirmed by statistical analyses considering a Monte Carlo-based approach. It is also shown that more realistic flashover rate estimates can be obtained in the statistical analysis of lightning-induced voltages provided a reduced threshold level (1.2 CFO in the particular case evaluated in this paper) is considered instead of the 1.5 CFO level traditionally used in this type of study.