Study region Southeastern United States Study focus The objective was to evaluate the ability of two modified SCS-CN models to predict direct runoff (DRO) and peak discharge rate (Qp) for selected storm events in three forested watersheds in the region - one low-gradient system in South Carolina, two high-gradient upland systems in North Carolina, and a mid-gradient upland system in Arkansas. New hydrological insights for the region The calculated peak discharge rate Qp values by all methods were unsatisfactory when using the default pond and swamp adjustment factor (Fp) value of 0.72 recommended in the SCS TR55 guideline, indicating that use of the default Fp value may result in erroneous Qp estimates for forest watersheds with high retention capacities. These findings, indicating the superiority of the modified Sahu-Mishra-Eldo (SME_m) method to the original SCS method in runoff calculations and substantially lower Fp values in the associated Qp method, are significant for hydrologists and engineers who frequently apply the methods in design of storm water management structures including road culverts on forested landscapes.