Forecasting plays a vital role in tourism planning and tourism promotion strategies requiring an estimate of future tourism demand to fulfill national aspirations. This article investigates the applications of time series forecasting techniques namely exponential smoothing and univariate Autoregressive Integrated Moving Average (ARIMA) models to predict the number of arrivals from different countries to the state of Penang in Malaysia by air and sea, based on available data from January 2002 to December 2007. The month-wise seasonal factor has been determined and consequently the peak, moderate and lean months in terms of international tourist arrivals to Penang has been detected. Further, it is worthwhile to study the cluster of countries visiting Penang to identify the arrival pattern from different origins.