Recent developments in the U.S. pecan industry appear to limit the utility of past research. The importance of pecan variety has emerged as an issue which could alter past results. The linear and double-log models previously fitted to all pecans (averaged) data may be too restrictive and hence, are less useful for variety-specific analysis. Past research also analyzed price turning points using nominal data. This study investigated functional form and data-averaging problems by fitting separate flexible Box-Cox price-dependent models for all-pecans and each variety of pecans (1970/71 -1988/89 deflated data). Results indicate: other nuts substitute for different pecan varieties, estimated all-pecans price flexibility is biased and clouds variety-specific flexibilities, and restrictive functional forms are inappropriate.