This study assessed the effects of polymorphic variants of gutathione-S-transferase and metallothioneins on profiles of urinary arsenic species. Drinking groundwater from Margarita and San Fernando, Colombia were analyzed and the lifetime average daily dose (LADD) of arsenic was determined. Specific surveys were applied to collect demographic information and other exposure factors. In addition, GSTT1-null, GSTM1-null, GSTP1-rs1695 and MT-2A-rs28366003 genetic polymorphisms were evaluated, either by direct PCR or PCR-RFLP. Urinary speciated arsenic concentrations were determined by HPLC-HG-AFS for species such as As-III, As-V, monomethylarsonic acid (MMA), dimethylarsinic acid (DMA), and total urinary As (TuAs). Primary methylation index (PMI) and secondary methylation index (SMI) were also calculated as indicators of the metabolic capacity. Polymorphisms effects were tested using multivariate analysis, adjusted by potential confounders. The As concentrations in groundwater were on average 34.6 +/- 24.7 mu g/L greater than the WHO guideline for As (10 mu g/L). There was a correlation between As concentrations in groundwater and TuAs (r = 0.59; p = 0.000). Urinary inorganic arsenic (%InAs) was associated with GSTP1, LADD, GSTP1*Age, GSTP1*alcohol consumption (r(2) = 0.43; likelihood-ratio test, p = 0.000). PMI was associated with sex (r(2) = 0.20; likelihood-ratio test, p = 0.007). GSTPI (AG + GG) homozygotes/heterozygotes could increase urinary %InAs and decrease the PMI ratio in people exposed to low and high As from drinking groundwater. Therefore, the explanatory models showed the participation of some covariates that could influence the effects of the polymorphisms on these exposure biomarkers to As. (C) 2018 Elsevier Ltd. All rights reserved.