Electronic somatic cell counts (ESCC) and relative cell volume distributions (RCVD) in milk samples were obtained simultaneously on a Coulter Counter Model TA and their potential assessed, either singly or in combination, for identifying mastitic quarters and cows in 2 herds. Using log (ESCC) alone the probability P m of misclassifying quarters infected with major pathogens was 6-7 % for herd 1 and 18-1% for herd 2. The inclusion of RCVD variables with log (ESCC) did not substantially improve the chance of detecting quarters infected with major pathogens. However, the RCVD variables did improve the chance of identifying quarters infected by major and minor pathogens compared to the use of log (ESCC) alone, the P m figures being reduced by about one third.When testing composite quarter samples, where milk from infected quarters was diluted with that from uninfected quarters, the inclusion of RCVD variables decreased the P m compared to the use of log (ESCC) alone. In herd 1 the improvement was from 26-1 to 3*0 %, while in herd 2 the effect was less marked, the P m being reduced from 28-6 to 21-6%. However, as the RCVD assessment is provided simultaneously with ESCC, the additional information is obtained at minimal cost.Electronic somatic cell counts (ESCC) of bulk milk are considered the most efficient means of ranking herds for the prevalence of sub-clinical mastitis (Reichmuth, 1975;Westgarth, 1975). Newbould (1975) claimed that electronic analysis of cell volume distributions using the Coulter Counter Model TA assisted in the interpretation of ESCC, and gave examples using mastitic and non-mastitic quarter milk, bulk milk and colostrum.Hoare et al. (1975) and Sheldrake et al. (1977) reported that cell volume analysis using the Coulter C1000 Channelyzer aided in detecting mastitis in both quarter and herd milk. Newbould (1978), again using the Model TA, gave data on cell count and one channel of the cell volume histogram of milk samples from quarters of known infection status. He attempted to compare the usefulness of cell volume analysis with cell count by comparing the frequency with which bulk milk samples exceeded arbitrarily selected limits for cell count and one channel of the cell volume histogram. This paper reports the use of the Coulter Model TA for examining the cell volume histogram of milk samples from quarters and cows of known infection status. Three cell volume channels of the histogram were assessed, alone or in combination with