The authors examine the supply response of the Greek broiler market. A generalized autoregressive conditional heteroskedasticity (GARCH) process is used to estimate expected price and price volatility; price and supply equations are estimated jointly. In addition to the standard GARCH model, several different symmetric, asymmetric, and nonlinear GARCH models are estimated. These models use different conditional variance specifications (linear or nonlinear) to grasp some additional empirical regularity of data like asymmetry. Asymmetric price volatility means that different volatility is recorded in the case of a fall in prices than an increase in prices by the same amount. The possible existence of asymmetry in the producer's price volatility gives useful information about market structure and possible market power. The empirical results indicate that among the estimated GARCH models the nonlinear asymmetric GARCH model (NAGARCH) seems to better describe producers' price volatility of the Greek broiler industry. Furthermore, the empirical findings show that price volatility is an important risk factor and broiler feed price is the most significant cost factor of the supply response function. Finally, the model provides forecasts for quantity supplied, producers' price, and price volatility. [EconLit. Classifications: Q110, C510, D200]. © 2010 Wiley Periodicals, Inc.