The objective of this study was to assess the effectiveness of a two-year education program on changing knowledge, attitudes (AT), subjective norms (SN), and intention (I) to consume a low-fat diet. Pre-menopausal women (n=278) were recruited through Cooperative Extension Service (CES) at seven Illinois sites. Approximately half served as controls receiving no education. The intervention consisted of six monthly education sessions, six months of fact sheet mailings, and four educational follow-up sessions. Knowledge, AT, SN, and I were measured at baseline, 6, 12, and 24 months. Knowledge was measured using the CES Impact Indicators instrument; AT, SN, and I instruments were developed using the Theory of Reasoned Action (TRA). T-tests were used to measure differences between treatments (T) and controls (C) and analysis of variance for differences among the time points. At baseline, there was no difference between T and C for knowledge, AT, SN, or I to consume a low-fat diet. There was no significant difference between T and C at later time periods for AT scores because scores were positive and remained so. For I, there was a significant difference between groups at six months (p=0.05) and 24 months (p=0.048), but I was strong at each time period for both T and C. For SN, there were no significant differences between T and C for any time period. Within the T group, education resulted in no significant difference at any time period for any variable except knowledge (p=0.000). None of the TRA or knowledge variables were correlated with body weight, fat, or waist:hip ratios, indicating that those more at CVD risk did not know more or have a different attitude or intention about their diet than those at lower risk. Regression analysis of TRA variables resulted in significant equations (p=<0.012) predicting intention to consume a low-fat diet for both T and C at each time period.