The natural distribution of bacteria in films prepared for microscopic counting is not taken into consideration when square films are used or random observations are made. Prepn. of circular films and microscopic sampling along 2 diams. at rt. angles takes into account the concentric distr. of bacteria and minimizes the effect of abnormal distribution. For reliable counts a suitable fixative must be used; this is discussed. Attempts were made to learn the minimal number of fields permissible to sample in order to obtain estimates of bact. numbers substantiated by counting a larger number of fields. The relation between the avg. number of bacteria per field and the number of fields is discussed.