of 0.01-50% v/v HCl, its sorption onto the surface of the PTFE fiber in the microcolumn, elution with 3% m/v KBH4 in 0.5% m/v KOH, hydride generation with 15% v/v HCl solution and determination with atomic fluorescence spectrometry. The alkaline KBH4 solution was employed not only as an efficient eluent but also as the reductant to produce selenide anion for subsequent hydride generation. With a consumption of 12.4ml of sample solution, a detection limit (3δ) of 4ng l-1 was achieved at a sample throughput of 26 samples h-1. The retention efficiency was 67%, being five times that obtained with a PTFE knotted reactor (KR). The linear ranges of the diagrams of signal intensity against sample loading flow rate, and sample loading time, were much wider than those achieved with a KR, offering more potential for obtaining high sensitivity by increasing sample loading rates and/or sample loading time. The precision for 11 replicate measurements of 0.5µg l-1 Se (IV) was 1.5% (RSD). The recoveries of Se (IV) from natural water samples varied from 96 to 108%. The concentration of Se (IV) in a certified reference material (NASS-5, sea-water) determined by the present method with a simple aqueous standard curve calibration was in good agreement with the certified value. The developed method was also successfully applied to the detection of Se (IV) in local river water, lake water and sea-water samples.