A technique for determining the speciation of chromium in natural waters is presented to distinguish Cr(VI), active inorganic Cr(M) and non-active organic Cr(IH) complexes. The technique is based on catalytic cathodic stripping voltammetry with adsorption of Cr(III)-DTPA complexes (CCSV-DTPA). Organic-Cr(III) complexes are not electrochemically active, whereas free aqua Cr3+ and its hydroxyl ions active. According to different behaviors of Cr(VI) and active Cr(HI) in buck solution and at electrode interface, a new calibration for [Cr(III)](active), and an entire protocol for Cr speciation are proposed. Peak currents for active Cr(III) decreased and could be fitted exponentially. Following the fitting curve for first 10-20 min, the extrapolated peak current at time zero (i(pt0)) was used as a measure of [Cr(III)](active) [Cr(Vl)] was measured 60 min after DTPA addition with a stable peak current when Cr(III)-DTPA complexes were all converted to a non-active species. The operational conditions for CCSV-DTPA were optimized. The detection limit and the recovery of [Cr(Vl)] and [Cr(III)](active) were, respectively, 0.1 nM and (100 +/- 3)% at 10 nM level in the presence of organic ligands. The speciation procedure was applied to natural samples collected from Swiss rivers, lakes, soil drainage, and landfill leachate. The [Cr](total) measured by GF-AAS and by CCSV-DTPA is close in nanomolar. The results indicate occurrence of non-active organic-Cr(M) complexes in natural waters. The fractions of [Cr(III)] were substantial in some samples from pollutant small rivers and soil leachates. Further application of the technique with evaluation of species alteration would facilitate studies related to chromium transportation, transformation and remediation in natural and contaminated environments. (C) 2001 Elsevier Science B.V. All rights reserved.