Results confirm most other research on milk composition and yield. All responses were affected by climate, some considerably more than others, if percent of error variance is the criterion. Jersey yields were less sensitive to climate than were Holstein, but Jersey milk composition appeared more sensitive. Somatic cell count (REF), a measure of mastitis, was affected by climate but less than all other variables except protein/fat and LM%. Needed are estimates of interactions between climatic effects and response surface models to quantify possible improvement in performance following environmental modification. Genetic correlations between milk yield and REF and chloride % suggest that single trait selection for milk yield might increase incidence of mastitis although phenotypic correlations indicate that high yields and absence of mastitis are correlated.