The International Agency for Research on Cancer determined that bromate is a potential human carcinogen, even at low micro/l levels in drinking water. Bromate is commonly produced from the ozonation of source water containing naturally occurring bromide. Traditionally, trace concentrations of bromate and other oxyhalides in environmental waters have been determined by anion exchange chromatography with an IonPac AS9-HC column using a carbonate eluent and suppressed conductivity detection, as described in EPA Method 300.1 B. However, a hydroxide eluent has lower suppressed background conductivity and lower noise compared to a carbonate eluent and this can reduce the detection limit and practical quantitation limit for bromate. In this paper, we examine the effect of using an electrolytically generated hydroxide eluent combined with a novel hydroxide-selective anion exchange column for the determination of disinfection byproduct anions and bromide in municipal and bottled drinking water samples. EPA Methods 300.1 B and 317.0 were used as test criteria to evaluate the new anion exchange column. The combination of a hydroxide eluent with a high capacity hydroxide-selective column allowed sub-microg/l detection limits for chlorite, bromate, chlorate, and bromide with a practical quantitation limit of 1 microg/l bromate using suppressed conductivity detection and 0.5 microg/l using postcolumn addition of o-dianisidine followed by visible detection. The linearity, method detection limits, robustness, and accuracy of the methods for spiked municipal and bottled water samples will be discussed.