The association between arsenic (As) exposure and diabetes is not clearly defined for populations exposed to low or moderate levels of inorganic As (iAs) in drinking water (< 150 μg/L). In the present study, the relationship between iAs concentration in drinking water (contaminated at a median level of 10.5 μg/L) or As biomarkers (ie, urine and nails) and diabetes or prediabetes (defined as level of glycosylated hemoglobin - HbA1c – higher than 6%, self-reported diagnosis of diabetes by a physician, or the use of insulin or oral hypoglycemic drugs) was evaluated in 257 adults from Canada. For that we used logistic regression models and reported the odds ratio (OR) comparing participants in the 80th vs 20th percentile of iAs exposure indicators. The association between iAs exposure indicators and HbA1c was also explored for 234 adults and 35 children not taking insulin or oral hypoglycemic drugs using a linear regression analysis. All models were controlled for confounding variables (age, gender, first-degree family history of diabetes, obesity or overweight in adults’ model). We attempted to exclude adults with organic arsenic of marine origin in their urine by removing participants with detectable arsenobetaine or arsenocholine in urinary models. iAs biomarkers (toenail and urine) were not associated with diabetes or prediabetes in adults. iAs in well water was associated with a borderline significantly increased odds of diabetes or prediabetes (OR = 2.39; 95% CI: 0.99–5.72). Higher well water iAs concentrations were significantly associated with increased HbA1c in both adults and children (β: 0.002; p = 0.041 and β: 0.003; p < 0.0001 respectively). In children, HbA1c was also associated with toenail As concentration (β: 0.18; p = 0.016). These results suggest low-level iAs exposure is associated with a continuum of dysglycemia.