A stochastic approach to characterization of temporal rainfall patterns is proposed wherein a rainfall event is defined as an unbroken sequence of consecutive daily rainfalls. A stochastic model is developed to determine the probability distribution of cumulative rainfall amounts at the end of each day within a total rainfall event duration of n days. The model is structured such that the distribution of daily rainfall depths can be approximated by an exponential distribution and the rainfall occurrence process can be described by a first-order stationary Markov chain. An illustrative example was presented, using a 32-year daily rainfall record at Dorval Airport on Montreal Island. The results of this example have demonstrated the adequacy and descriptive capabilities of the model. It can be concluded that the methodology proposed here seems to be more general and more flexible than those that have been used in previous investigations. Key words: daily rainfall process, temporal rainfall pattern, stochastic model, stochastic hydrology, exponential distribution, Markov chain.