Pluviographic data at 15 min interval from 6 stations in Pulau Penang of Peninsular Malaysia were used to compute rainfall erosivity factor (R) for the revised universal soil loss equation (RUSLE). Three different modelling procedures were applied for the estimation of monthly rainfall erosivity (EI30) values. While storm rainfall (P) and duration (D) data were used in the first approach, the second approach used monthly rainfall for days with rainfall � 10 mm (rain10) and monthly number of days with rainfall � 10 (days10). The third approach however used the Fournier index as the independent variable. Based on the root mean squared error (RMSE) and the percentage error (PE) criteria, models developed using the Fournier index approach was adjudged the best with an average PE value of 0.92 and an average RMSE value of 164.6. Further, this approach was extended to the development of a regional model. Using data from additional sixteen stations and the Fournier index based regional model, EI30 values were computed for each month. ArcView GIS was used to generate monthly maps of EI30 values and also annual rainfall erosivity (R). The rainfall erosivity factor (R) in the region was estimated to vary from 9000 to 14,000 MJ mm ha- 1 h- 1 year- 1. © 2007 Elsevier B.V. All rights reserved.