Potentially toxic elements (PTEs) in mining-impacted wetlands can transfer from soil and water to forage and grazing livestock, resulting in dietary exposure for nearby communities. In this study, arsenic (As), lead (Pb), and cadmium (Cd) were quantified in key environmental matrices (soil, surface water, and forage) and in sheep tissues (liver, kidney, and muscle) from six georeferenced grazing sites in the Ite coastal wetland (Tacna, Peru) during the dry season. Samples were acid-digested following U.S. EPA protocols and analyzed by atomic absorption spectrometry (AAS) under QA/QC procedures (certified reference materials, blanks, duplicates, and matrix spikes); matrix-specific detection and quantification limits are reported. Arsenic dominated the contamination profile (forage: 428.6 mg kg⁻¹, dry weight; soil: 48.61 mg kg⁻¹; water: 0.97 mg L⁻¹) and was detected in sheep tissues (kidney: 0.1577 mg kg⁻¹; muscle: 0.1538 mg kg⁻¹; liver: 0.0644 mg kg⁻¹). Lead and cadmium were <LOQ in muscle and liver but were measurable in kidney (Pb: 0.0415 mg kg⁻¹; Cd: 0.0011 mg kg⁻¹). To support the interpretation of trophic transfer, screening transfer/bioaccumulation metrics (soil-to-forage, forage-to-tissue, and water-to-tissue) were calculated. Human dietary exposure was screened using estimated daily intake (EDI), hazard quotient (HQ), and margin of exposure (MOE). Because arsenic speciation was not measured, inorganic arsenic (iAs) risk was evaluated as a conservative upper bound (100% iAs) and through sensitivity scenarios (10–50% iAs). Under a high-meat-consumption scenario (300 g day⁻¹), the upper-bound assumption yielded HQ = 2.2 and MOE = 0.46; however, scenario analyses indicate that risk conclusions are highly dependent on the assumed iAs fraction. Overall, the results identify arsenic as a priority contaminant and support targeted grazing management, the provision of low-arsenic water sources, and remediation and monitoring actions to reduce exposure in vulnerable rural communities.