1.Three classes of immunoglobulins have been identified in lacteal secretions. IgG1 appears to be selectively transferred through the glandular epithelium into the udder. Presumably any method of immunization that stimulates circulating antibodies of the IgG1 class will be beneficial, resulting in IgG1 antibodies in the milk.2.IgA class immunoglobulins are also found in the milk, however, in lower concentration than IgG1 immunoglobulins. IgA is produced locally in the gland since an increase is seen in antigen-infused quarters over noninfused quarters. A factor that might be important is that probably the total amount of any immunoglobulin in lacteal secretions will not be increased by systemic immunization whereas local immunization does increase the amount of IgA.3.The method and route of immunization is possibly very important. Intramuscular, subcutaneous, or other routes that result in systemic stimulation of antibodies, to be effective, must for the most part stimulate class IgG1 immunoglobulins. Hammer, Kickhofen, and Henning (7) have recently published that there is a quantitative shift of antibody activity from IgG1 to IgG2 between the first and third courses of immunization of cattle (7). These IgG2 antibodies would be of little value for mammary gland immunity insofar as antibodies in the milk are concerned. However, since the tissues of the mammary gland are involved in a mastitis infection, specific antibodies of all immunoglobulin classes may be beneficial. Immunization by inoculation of a Streptococcus agalactiae, type II preparation into the area of the supramammary lymph node stimulates protective antibodies in the milk. These antibodies have been shown to be of IgG class principally, with less IgA. Finally, by intramammary infusion of antigen into one or more quarters of a nonlactating gland, an increase in specific antibodies of only the IgA class was noted.4.Progress in the determination of the immunogenic substance necessary to stimulate resistance to staphylococcal mastitis has been slow. Serological typing of staphylococci is not available; however, on the basis of phage typing, most mastitis staphylococci may fall into a dozen or so types (14). Serological typing and the immunogenic response of the group B streptococci are much further advanced. The immunology of the group C and group E streptococci is also in a primitive state. It probably is not necessary to separate and purify each immunogen that is required in a preparation; however, it is necessary that inactivated bacteria of each different antigenic type be present in a bacterin in order to stimulate all the various antibodies necessary.One can speculate that a combination of immunization routes that result in the stimulation of several of the classes of immunoglobulins might produce sufficient protective antibodies in the milk and in the mammary gland tissues to increase resistance. Local stimulation would increase specific IgA while systemic stimulation would result in specific IgG1, IgG2 and IgM. Local stimulation would presumably have to occur in the nonlactating gland with a nonirritating or nontissue damaging antigen. The antigen may be a bacterin consisting of a combination of each of the types of staphylococci and streptococci and be inactivated with formalin or some other inactivant. The bacterin should also contain staphylococcal toxoids.