Using evidence from 46 gauged small catchments in the United Kingdom, this paper demonstrates that the methods most commonly used for estimating design flows and greenfield run off rates on small catchments do not perform as well as alternative methods. Their results show larger error and a bias towards underestimation of the median annual flood. In contrast, newer methods from the Flood Estimation Handbook (FEH), when applied to small catchments, tend to have lower error and less bias. The paper investigates the theoretical and empirical support for four methods: Institute of Hydrology Report 124, Agricultural Development and Advisory Service (ADAS) Report 345, the FEH Statistical method and the Revitalised Flood Hydrograph method. It compares the results from all four methods with those from direct analysis of flood peak data and discusses the implications of scaling down flows to the development plot scale.