Railway formation (embankments and cuttings) in Queensland are experiencing increased maintenance costs due to rainfall induced erosion and sedimentation problems. The erosion problems can be reduced by establishing grass cover on the steep slopes (batters). The objective of the research reported in this paper was to develop a continuous rainfall-runoff-soil erosion model to investigate grass cover effects on runoffand soil loss on railway formation steep slopes. It utilizes the Saint Venant continuity and momentum equations for overland flow, and a modified Green-Ampt model for infiltration on steep slopes. The erosion component of the model consists of two different methods, based on (1) the modified universal soil loss equation, and (2) steep slope erosion dynamics. The efficiency of the model was evaluated by the percentage error and the Nash-Sutcliffe efficiency values. Field trials data gathered from 10-m wide sections with different grass cover percentages were used to calibrate the one-dimensional distributed rainfall-runoffmodel. Rainfall and runoffwere monitored at 1-min intervals while the soil loss data were collected at different sampling intervals. The model has successfully predicted runoffand soil loss from the plots for the majority of the cases, with Nash-Sutcliffe efficiency values varying between 0.43 and 0.99 for 0% grass cover plots, 0.06 and 0.97 for 50% grass cover plots, and -0.42 to 0.94 for 100% grass cover plots. The percentage error values for soil loss vary from 5 to 58% for 0% grass cover plots for the steep slope erosion dynamics method and 4-59% for the modified universal soil loss equation-based method. For 50% grass cover plots, the percentage error values vary from 2 to 160 and 18 to 86%, for the steep slope erosion dynamics and modified universal soil loss equation-based methods, respectively. With the aid of an antecedent moisture parameter which varies with grass cover percentage, continuous simulation of runoffand erosion can be carried out, using either long records of observed fine timescale rainfall or that derived from a stochastic rainfall model.