Flooding is a major problem in the UK as recent high profile events in the summers of 2006 and 2007 have shown. In these events the damage to property and belongings ran into billions of pounds and a number of people were injured or lost their lives in these events. Therefore, predicting the location and severity of flooding is extremely important in preventing these losses. Current computer models for predicting flooding are highly accurate, but take a very long time to run even on the fastest computers. This project intends to use a technique known as cellular automata, a model based on the localised interactions of small cells, to simulate flooding in such a way that it will be possible to run complicated scenarios on a standard PC. The new approach will gain efficiency by making use of the fact that each cell can only 'see' the cells closest to it and the project will investigate the best ways of allowing each cell to communicate with its neighbours. The approach will be tested over a number of different flooding scenarios and compared with existing methodologies to demonstrate its accuracy and increased efficiency over standard methods.