Bromide electrolysis was carried out on laboratory-scale cells in the range of 1-1,005 mg [Br-] dm(-3) using boron-doped diamond (BDD) anodes. These studies were part of fundamental research activities on drinking water electrolysis for disinfection. Synthetic water systems were mostly used in the experiments, which varied the temperature between 5 and 30 A degrees C, the current density between 50 and 700 A m(-2), and the rotation rate of the rotating anode between 100 and 500 rpm (laminar regime). Hypobromite and bromate were found as by-products, as expected. Bromite was not detected. Higher bromate levels were formed at higher current density, but no clear relationship was observed between bromate concentration and the rotation rate or temperatures between 5 and 30 A degrees C. Bromate yields higher than 90% were found at higher charge passed. Perbromate was found as a new potential synthesis or disinfection by-product (DBP), but no perbromate was detected at the lowest bromide concentrations and under drinking water conditions. The perbromate yield was about 1%, and somewhat lower when bromate was used as a starting material instead of bromide. At a temperature of 5 A degrees C more perbromate was detected compared with experiments at 20A degrees. Approximately 20 times more perchlorate was formed compared with perbromate formation in the presence of chloride ions of equimolar concentration. State of mechanistic considerations is presented and a mechanism for perbromate formation is proposed. The reaction from bromate to perbromate was found to be limited that is in contrast to the earlier studied chlorate-to-perchlorate conversion. In the measured concentration range, reduction processes at the mixed oxide cathode showed a much higher impact on the resulting concentration for perbromate than for bromate.