14C-Labelled milk proteins and peptides were required for studies on the nitrogen nutrition and proteolytic enzymes of starter bacteria. Therefore, a cow was injected with a mixture of 14C-labelled amino acids (2mCi) and milked at intervals over 28 h. The first milking (3.75 h) contained the most active components with the caseins, β-lactoglobulin and α-lactalbumin having a specific activity of 2.3 x 106 disintegrations per min (dpm) / g while lactose and triglycerides had specific activities of 2.3 x 105 dpm / g and 3.4 x 104 dpm/g, respectively. Thirteen amino acids isolated from acid hydrolyzed milk protein contained radioactivity. Only 6% of the label injected was recovered in milk protein in the first 28 h. The nitrogen nutrition of Streptococcus cremoris AM2 and E8 was studied by adding 14C-labelled milk proteins, peptides and single amino acids separately to unlabelled milk and then determining the incorporation of radioactivity into bacterial protein during growth. A specially prepared low heat skim milk powder having a content of potential nitrogen sources similar to that of fresh milk, was reconstituted for use as the growth medium. At low cell densities, free amino acids and peptides were used as nitrogen sources. As the cell density increased, milk protein became an increasingly important nitrogen source and the cells became dependent on their cell wall-bound proteinase. All caseins tested, βlactoglobulin and α-lactalbumin were used as nitrogen sources by both strains. A proteinase assay was evaluated in which 14C-labelled casein (2.3 x 106 dpm/g) was used as the substrate. This assay was used to study the spontaneous release of cell wall-associated proteinases from milk-grown cells of lactic streptococci. Eight strains of S. cremoris and two strains of S. lactis released proteinase when cells were held in buffer. An exception was S. cremoris ML1 which did not release significant activity. The rate of proteinase release increased with rise in temperature (0 to 34°C) and pH (5.5 to 8.7) although inactivation was apparent at 34°C and pH 8.7. With all strains, release of proteinase was suppressed by the addition of CaCl2 to the buffer, by lowering the temperature to 0°C, or by lowering the pH to 5.5. The rate of proteinase release varied markedly with different strains. A possible mechanism for the release of proteinase from the cell wall is discussed. Cheddar cheese was made with cultures containing different proportions of proteinase-positive (Prt+) and proteinase-negative (Prt-) cells. This allowed the level of starter proteinase to be varied while the total concentration of starter cells in the curd at salting was kept constant. Cheeses with 45 to 75% Prt- cells developed significantly less bitterness than cheeses containing only Prt+ cells, thus providing direct evidence that the level of starter proteinase has a role in bitterness development in Cheddar cheese. The involvement of starter peptidases in the removal of bitter peptides is discussed.