Ruminant feed supplements are price sensitive and are effected by the continuous fluctuation of other raw material feed prices. Therefore, improving the efficiency of production and reducing cost of supplement concentrates for dairy cows are becoming increasingly important both for the smallholder and commercial dairy farmer. This can be overcome by replacing expensive energy and protein feeds with cheaper by-products. During periods of high maize prices, replacing maize with lower cost high fibre by-products becomes an economically viable option. Palm kernel expeller (PKE) fits the profile of a low cost, high fibre by-product. The aim of this study was to determine the effect of different inclusion levels of PKE in dairy concentrates for Jersey cows on milk production, milk composition, body weight (BW) and body condition score (BCS) change, rumen parameters and in situ ruminal kikuyu/ryegrass pasture degradability of dry matter (DMd) and neutral detergent fibre (NDFd) as well as NDFd rate (NDF kd). The study was conducted at the Outeniqua Research Farm situated near George in the Western Cape and cows grazed high quality kikuyu/ryegrass pasture during spring. Forty eight multiparous high producing Jersey cows were blocked according to 4% fat corrected milk (FCM), days in milk (DIM) and lactation number and randomly allocated to three treatments (control, low PKE, and high PKE). The PKE inclusion level in the control, low PKE, and high PKE treatment concentrates was 0, 20, and 40%, respectively. The PKE replaced part of the maize and protein sources in the concentrate. Milk yield was recorded daily and milk composition was determined in two week intervals over a 60 d period, after a 21 d adaptation period. Additionally, eight lactating rumen-fistulated cows were randomly allocated to the control and high PKE treatment in a two period crossover design. Ruminal pH, volatile fatty acids (VFA’s), ruminal ammonia-nitrogen (NH3-N), and in situ ruminal kikuyu/ryegrass pasture DMd, NDFd and NDF kd were measured. Cows received 6 kg (as is) concentrate per day divided over two milking periods and strip grazed kikuyu/ryegrass pasture as one group. Milk yield and milk fat content did not differ (P > 0.05) between treatments and were 21.3, 21.3 and 20.7 kg/cow/d and 4.63, 4.65, and 4.66% for cows receiving the control, low PKE and high PKE treatments, respectively. Milk protein, milk urea nitrogen (MUN), BW and BCS did not differ (P > 0.05) between treatments. Total VFA’s, average ruminal pH, ruminal NH3-N, and in situ ruminal kikuyu/ryegrass pasture DMd and NDFd as well as NDF kd did not differ (P > 0.05) between treatments. The acetic to propionic acid ratio was, however, higher (P < 0.05) for cows supplemented with the high PKE treatment. It can be concluded that partial replacement of maize with 20 or 40% PKE in a lactating dairy cow concentrate did not affect milk yield, milk fat content, milk protein content, somatic cell count (SCC), BW, or BCS. Rumen fermentation was unaffected and a healthy rumen environment was sustained. The replacement of higher cost maize and soybean oilcake by a lower cost PKE decreased feed cost. It is however not recommended to include PKE at 40% in the concentrate due to the increased time spent by cows in the milking parlour and the low palatability of PKE, which could lead to the tendency of increased concentrate refusals. It can be extrapolated from the data obtained from this study that milk production will be sustained when PKE is fed to cows on pasture at 2.4 kg/cow/day. / Dissertation (MSc Agric)--University of Pretoria, 2013. / gm2014 / Animal and Wildlife Sciences / unrestricted