Women of African descent have a high prevalence of diseases caused by insulin resistance. To positively impact cardiometabolic health in Black women, effective screening tests for insulin resistance must be identified. Recently, the TG/HDL-C ratio has been recommended as a tool to predict insulin resistance in overweight people. While the ratio predicts insulin resistance in White women, it is ineffective in African American women. As there are no data for African women, we tested the ability of the TG/HDL-C ratio to predict insulin resistance in Black women from South Africa, West Africa and the United States. For comparison, the ratio was also tested in White women from South Africa. Participants were 801 women (157 Black South African, 382 African American, 119 West African, 143 White South African, age 36 +/- 9y [mean +/- SD]). Standardized scores were created from log-transformed homeostasis model assessment-insulin resistance values from each population. Participants in the upper third of their population distribution were classified as insulin-resistant. To predict insulin resistance by the TC/HDL-C ratio, area under the receiver operating characteristic (AUC-ROC) curve was used and criteria were: 0.50 for no discrimination and > or = 0.70 for acceptable. Seventy-one percent of the Black women were overweight vs 51% of White women (P<.01). In overweight White women, AUC-ROC curve for prediction of insulin resistance by TG/HDL-C was 0.76 +/- 0.06, but below the 0.70 threshold in each group of overweight Black women (Black South African: 0.64 +/- 0.06, African American: 0.66 +/- 0.03, and West African: 0.63 +/- 0.07). Therefore, TG/HDL-C does not predict insulin resistance in overweight African American women and this investigation extends that finding to overweight Black South African and West African women. Resources to identify effective markers of insulin resistance are needed to improve cardiometabolic health in women of African descent.