Mastitis is inflammation of the mammary gland caused by a pathogenic microorganism, although the organism may not be present at the time of examination. Clinical mastitis is recognized by abnormal milk, gland swelling, and/or illness of the cow. Subclinical mastitis is characterized by apparently normal milk, with an increase in somatic cell count (SCC) due to the influx of leukocytes. The SCC provides a good quantitative estimate of the degree of inflammation. Subclinical mastitis can also be characterized by culturing the milk of cows with a high SCC or all cows in the herd. A realistic goal for an acceptable prevalence of mastitis in a herd is determined by a number of factors including the extent of infection currently present, the type of infections, and the control measures used to reduce infections in that herd. The prevalence of subclinical mastitis in dairy herds is often surprising to producers. Identifying infection by milk culture can be a good aid in controlling a high SCC in order to avoid regulatory penalties or to maintain a low SCC for quality premium payments. Herd outbreaks of clinical mastitis 'or cases in individual cows, especially highly valued cows, are of interest. Designing a management strategy to reduce new cases and improve prognosis for cows already affected is aided by identifying the etiologic agents of mastitis. Knowledge of antibiotic susceptibility or resistance of mastitis pathogens is valuable to veterinarians and dairy producers. The identification of pathogens along with their antibiogram results should help in developing therapy protocols for dairy farms. The Laboratory Handbook on Bovine Mastitis published by the National Mastitis Council [1] is an excellent laboratory manual and useful in any veterinary practice. Many farms, however, are large enough to employ professional staff capable of setting up and maintaining a milk-culturing laboratory. This "on-farm" diagnosis can be used in developing farm therapy protocols and monitoring the farm's mastitis control programs. Culture data provides additional information that can be used with routine SCC tests to determine the source or sources of new infections, as well as influence therapy decisions and appropriate use of antiobitics. Until better cow-side diagnostic tests are available, selective media used in diagnostic culture protocols provides a reasonable method for determining the major bacteriologic causes of mastitis. Under the supervision and direction of the herd or consulting veterinarian, accurate diagnosis can be made. Bacteria isolated on site that are not easily identified can be submitted to a diagnostic laboratory for identification or confirmation of the preliminary findings. The critical key to a useful on-farm laboratory is the proper training of personnel to collect appropriate samples and carefully handle them so that overgrowth of contaminates does not occur. Procedures in the on-farm laboratory need routine supervision by the farm veterinarian to monitor the quality of data and to appropriately use the information collected. This article addresses these issues that are essential in making good use of the diagnostic procedures to improve udder health on dairies.