The problems of sampling and of the statistical evaluation of microbial numbers in mineral waters.

Problems in sampling are considered and the implications of various sampling schemes discussed in terms of the risks involved to the users of the schemes. It is concluded that microbiological testing alone cannot guarantee the "safety" of a product. The reliability of analytical procedures of use in the microbiological examination of water is examined and a method of determining the state of statistical control of the procedure described. Finally control charts available for monitoring a manufacturing process are discussed with particular emphasis on the cumulative sum test and its application to colony counts from water.