This paper examines relationships between body weight, religion, and gender while controlling for relevant covariates and body mass index (BMI), a measure of physical/biological body type. Using data from the 2004 Survey of Texas Adults (n = 1,504), we present results of ordered logistic regression models which indicate that religious factors work distinctly for men and women when controlling for BMI. While church attendance is associated with lower odds of overweight perceptions among women, it is religious salience that is associated with lower odds of self-reported excess weight in men. Implications for research which associates religious and physiological factors are discussed.