Post-milking teat disinfection is an effective management practice in the reduction of the prevalence and incidence of mastitis. However, dipping and also spraying is time and labour consuming and may reduce parlour efficiency. A comparison was made between an automatic dipping and flushing (ADF) system and a reference manual dipping system. For two times 6 days, teats of 6 cows were disinfected using the ADF-system while teats of 6 other cows were dipped manually in a cross-over design. Milk samples for iodine content were collected 16 times. Vacuum drop and vacuum fluctuation in the short milk tube were estimated in the WUR-ASG milk recording laboratory using water instead of milk (vacuum level 40 kPa). In a challenge test the cluster flushing was evaluated by using an inoculum of Streptococcus agalactiae mixed with pasteurised milk. No difference between ADF system and reference manual dipping system were found in: milk yield, milk duration, average and maximum milk flow, average vacuum and cyclic vacuum variation. Differences between the ADF system and manual dipping were found for teat coverage (4 level visual scoring system): 2.92 versus 3.80 and the iodine content of the milk 155 versus 111 mu g/l. It was, however, far below the regulatory limit of 500 mu g/l. When challenged with an inoculum of Streptococcus agalactiae in pasteurised milk on the inside of the liners, the ADF system reduced Streptococcus agalactiae by 99.9% compared to no cleaning.