Various exponential smoothing models are estimated over the period. 1975-19
99 to forecast quarterly tourist arrivals to Australia from Hong Kong, Mala
ysia, and Singapore. The root mean squared error criterion is used as a mea
sure of forecast accuracy. Prior to obtaining the one-quarter-ahead forecas
ts for the period 1998 to 2000, the individual arrival series are tested fo
r unit roots to distinguish between stationary and non-stationary time seri
es arrivals. The Holt-Winters Additive and Multiplicative Seasonal models o
utperform the Single, Double, and the Holt-Winters Non-Seasonal Exponential
Smoothing models in forecasting. It is also found that forecasting the fir
st differences of tourist arrivals performs worse than forecasting its vari
ous levels. .