A novel, miniaturized micro-sequential injection Lab-on-Valve (muSI-LOV) system hyphenated with electrothermal atomic absorption spectrometry (ETAAS) is proposed for the automatic preconcentration and speciation analysis of Cr(III) and Cr(VI) utilizing solid-phase extraction on hydrophilic chelating Sepharose beads in the renewable bead injection (BI) mode. Exploiting on-line reduction of Cr(VI) to Cr(III), the aspirated sample solution is initially divided into two portions, which are treated simultaneously. Thus, while Cr(III) ions are separated from the matrix constituents and preconcentrated on the beads and subsequently eluted by a small volume of eluent (0.1 mol L-1 HNO3) and quantified by ETAAS, the Cr(VI) ions in the second portion are mixed with a reducing agent and parked under stopped-flow conditions in an open tubular reactor attached to one of the peripheral ports of the LOV unit. Following quantification of the native Cr(III) content, the Cr(III) generated from Cr(VI) plus the original Cr(III) are subjected to the same separation/preconcentration/elution procedure. All sample manipulations are controlled automatically by the integrated software. Under optimized chemical and physical conditions, the flow system, by using a total sample loading volume of 3.6 mL, featured retention efficiencies for Cr(III) as high as 86%, and preconcentration factors of 62 and 42 and detection limits (3s) of 0.010 and 0.020 mug L-1 for Cr(III) and Cr(VI), respectively, The relative standard deviations were 4.7 and 4.5% (n=6) at the 0.2 mug L-1 for Cr(III) and Cr(VI), respectively, when employing the microcolumn in a renewable fashion, while permanently used sorbent reactors yielded repeatabilities better than 3.0%. The proposed muSI-BI-LOV analyser was successfully applied to the speciation and determination of trace levels of Cr(III) and Cr(VI) in environmental samples. The method was validated by determination of chromium species in CRM and NIST standard reference materials, and by spike recoveries of surface waters. Statistical comparison of means between experimental results and the total chromium certified values for the CRM and NIST materials revealed the non-existence of significant differences at a 95% confidence level.