We assessed the threshold of homeostasis model assessment (HOMA) values to determine insulin resistance (IR) in a sample of Iranians and examined the associations of IR with metabolic syndrome (MetS). Study comprised 1327 non-diabetic and non-hypertensive subjects (438 men, 838 women; aged 20-77 years) incurred in four different locations (Tehran, Iran), 2005-2008. Lower limit of the top quintile of HOMA values in subjects without any metabolic abnormality was considered as the threshold of IR. This threshold was 1.8 (1.7 men; 1.8 women). Overall, 41.1% (36.3% men; 41.5% women) of subjects had IR. HOMA cut-off to determine MetS was 1.95 for ATPIII definition, and 1.85 for IDF. IR associated MetS (ATPIII: odds ratio (OR)=2.9, 95% CI=2.2-3.9, p<0.05; IDF: OR=2.94, 95% CI: 2.3-3.8, p<0.05). Sensitivity of MetS to detect IR was 22.4% for IDF and 36.2% for ATPIII criteria. In multivariate models, HOMA was predicted by waist circumference, and inversely by age and serum HDL-cholesterol. In this study, the cut point of HOMA to detect IR was lower than other populations. IR is an unyielding correlate of MetS; but definitions of MetS are insensitive measures of IR in our population.