Objectives To validate the veracity and stability of an automatic colony counts analyzer, and to study the feasibility of using the instrument in routine works.

Methods Twenty eight drinking water samples were randomly selected, and the number of colony was counted both by instrument method and artificial method. The number of colonies of one randomly selected water sample was counted for 10 times by the instrument method, and the results were compared with artificial method and statistically analyzed.

Results There was a high correlation (r equals to 0.976, P equals to 0.000) of the colony counts between instrument method and artificial method, and no significant difference on the means (t equals to 1.164, P equals to 0.254) of these two methods. The coefficient of variation (CV) for 10 times of determination of the same sample was 4.07%.

Conclusions The results showed that no appreciable difference between instrument method and artificial method on colony counts in water sample. The advantage of the instrument method was high precision, good repeatability and high efficiency. It is practical of using this automatic colony counts analyzer in routine works of testing colony counts in drinking water.