The authors study the influence of black coffee beans in mixture with Brasilian coffee of soft taste. Increasing percentages of black beans were tried: 0, 1, 2, 5, 10, 15, 20, 30, 40 and 50%. Two experiments were carried out, in a balanced incomplete block design, with t = 10 treatments, k = 4 plots per block, r = 6 replications, b = 15 blocks, L = 2. Each plot had 3 standard cups and was tasted by 3 tasters, who gave their opinion for each cup. Data thus collected are, therefore, 1080 in number, that is, 540 for each of the two trials. But each trial had indeed only 60 plots, to every plot corresponding the average of 9 opinions (3 tasters x 3 cups). The experimental technique was that established by FAIRBANKS BARBOSA and others (1962), with the changes indicated. The experiments gave results in good agreement, which led to the following conclusions: Black coffee beans injure coffee taste, and their effect is approximately linear. Percentages of black beans from 10% upwards carry out a sensible injury to the tasting quality of coffee. The regression equation is: Y = 3.792 - 0.0379 X, where X is the percentage of black coffee beans, and Y is the score corresponding to tasting categories.