In recent years, price volatility is becoming increasingly relevant to producers and consumers in the saturated food markets amidst stiff completion and globalisation. The analysis of price volatility is necessary to develop bidding strategies or negotiation skills in order to maximize profit. The generalised autoregressive conditional heteroscedasticity (GARCH) regression model is used to forecast foodstuff prices in Ghana over the period 1970 to 2006. The data used are monthly wholesale prices for maize, millet, and rice obtained from the Ghana Ministry of Food and Agriculture. The empirical results reveal that foodstuff prices exhibit high volatility with continual increasing prices over the study period. The results of the out-sample forecast reveal that maize, millet and rice prices would increase by 23%, 11% and 10% respectively in the next month. The study recommends the provision of adequate storage facilities, and farmers’ market centres in the districts to stabilize food prices. The increases in food prices have implications for food and nutrition situation of the poor in Ghana.