In a 2 × 2 × 3 factorial design grazing experiment we investigated the effect of fertilizer (none or 240 kg N/ha), amounts of clover grass available (low or high) and type and level of daily supplementary feed for each cow (3·5 kg barley, 3·5 kg concentrate mixture rich in protein and fat, or both, 7 kg) on the protein composition and renneting properties of their milk. The experiment was carried out in two successive grazing seasons (years) and included a total of 79 Danish Holstein cows. The effect on milk protein composition was determined in both years whereas the effect on renneting properties was determined only in the second year. Fertilization of the clover grass significantly decreased total milk protein concentration (− 1·4 g/kg; P < 0·01) and tended also to decrease the relative proportion of whey protein N. Fertilization had no effect on renneting properties. Increased availability of clover grass significantly increased milk protein concentration (1 g/kg; P < 0·05) and resulted in significantly poorer renneting properties, that is increased clotting time (P < 0·01) and decreased coagulum development. These effects seemed to be mediated through an effect on the pH of the milk (+ 0·05; P < 0·05) as the effect was markedly reduced when statistical correction was made for the actual pH. Use of the protein- and fat-rich concentrate mix (3·5 kg) significantly reduced the total protein content of the milk (P < 0·05) and increased the proportion of non-protein N (NPN) in total N compared with use of the other supplementary feeds (P < 0·05). We found no effect on renneting properties of the different supplementary feeds. Throughout the grazing season and independent of the main treatments, the NPN proportion of milk N increased at the expense of casein N. At the same time, renneting properties became poorer, especially with high clover grass availability.