Two ultraviolet (UV) systems were evaluated for use in water‐cooling devices: in‐tank (8 W, held in the centre of a 2‐L chilled water tank) and in‐line (6 W, treating chilled water at the point of dispense). The efficacy of these devices, under conditions of continual dispense (2 L/min), against a range of bacteria commonly associated with drinking water, was investigated. A rate of 1 L/min and the impact of a periodic dispense pattern (machine was not dispensing for 1 min between sampling) was examined with Escherichia coli. At 2 L/min, both devices achieved a 3 log reduction. At 1 L/min, both devices achieved greater than 4 log reduction. During periodic dispense, the in‐line device achieved a 3 log reduction, while the in‐tank device achieved greater than a 4 log reduction. UV was effective at reducing levels of bacteria commonly associated with water to acceptable levels. Under conditions of periodic use, the in‐tank device showed greater potential.