Obesity being associated with diabetes ([ Bermudez and Tucker, 2001 ]; [ Berber et al., 2001 ]), cardiovascular disease ([ Rexrode et al., 2001 ]), hypertension ([ Beegom et al., 1995 ]), dyslipidemia ([ Okosun et al., 2000 ]; [ Berber et al., 2001 ]), and some cancers ([ Fryzek et al., 2005 ]) has many negative effects on health; it is also associated with increased mortality ([ Must and Lipman, 1999 ]). The prevalence of obesity has increased dramatically in industrialized and developed countries ([ Mokdad et al., 2000 ]; [ de Onis and Blossner, 2000 ]), and overweight and obesity having been reported by the World Health Organization to be an escalating epidemic, worldwide ([ World Health Organization, 1997 ]). The majority of studies have recommended waist circumference (WC) as a better indicator of abdominal obesity and a better predictor for diabetes than other anthropometric variables ([ Wei et al., 1997 ]; [ Janssen et al., 2002 ]). Others have shown that the waist-to-hip ratio (WHR) is the best predictive anthropometric variable for development of type 2 diabetes ([ Esmaillzadeh et al., 2004 ]; [ Snijder et al., 2004 ]). Recently, a few studies have recommended waist-to-height ratio (WHtR) as the best diabetes predictor ([ Hsieh and Yoshinaga, 1995 ]). Since it has been proven that the predictive power of anthropometric variables is population-dependent ([ Molarius and Seidell, 1998 ]) and varies between different ethnic groups ([ Gallagher et al., 1996 ]), and the prevalence of obesity has increased dramatically in industrialized and developed countries, as well as in Iran ([ Pishdad, 1996 ]; [ Sarraf-Zadegan et al., 1999 ]), the present study was designed to identify the best anthropometric variable among WC, WHR, and WHtR as compared to body mass index (BMI) for predicting diabetes risk in Iranian men.