Introduction: Since different populations have different eating habits and chronic diseases, their fat intake instructions vary. The aim of this study was to examine determinants of fat consumption based on health belief model (HBM) on women's in Yazd city in 2015.
Methods: This cross-sectional study conducted on women in Yazd. The sample size of 203 women who were covered by health centers were selected randomly in two stages. Data were collected by two questionnaires. The first one was a researcher-made questionnaire consisting of: demographic questions and questions related to health belief model constructs, whose validity and reliability were proven. The second questionnaire was a revised version of the 168 item- food frequency questionnaire for the study of Tehran Lipid and Glucose which based on the taste of Yazd people, 10 food items were added. Data analysis was done by using SPSS software using central and spatial indices, Spearman correlation coefficient, linear regression and ANOVA.
Results: The mean score of the HBM constructs including perceived sensitivity, perceived intensity, perceived benefits, perceived barriers, cues to  action and self- efficacy were 4.22±14.93   (range 0-32), 3.07±12.02  (range 0-24), 1.79 ± 17.9 (range 0-28), 5.44 ±18.32  (range0 -32),  1.46±3.06   (range0 -11), 4.93±23.16   (range 0-32), respectively. The average daily dietary fat intake and the amount of received calories were 48.62±97.25  grams and 1156.8±2548.61 (Kcal), respectively. Percentage of solid and liquid oil consumption of all consumed oils were21.92±33.94%  and 38.5 ±54.33%, respectively. The cholesterol intake average was  149.33±261.42 mg. Based on Logistic regression analysis, none of the health belief model constructs were a strong predictor of fat intake.
Conclusion:    Due to the fact that health belief model constructs did not predict fat consumption behavior, it can be concluded that this pattern was not effective in explaining fat consumption behavior and other models can be used.

.