There are great variations in test systems used to measure pneumococcal opsonophagocytosis; therefore, in the present study, a detailed analysis of the human granulocyte opsonophagocytosis test with pneumococci was made. Variables affecting the test were evaluated. These were the amount of bacteria, the amount of antipneumococcal antibody and complement in the test. Correlation between phagocytosis percentage (percentage of cells taking part in the phagocytosis) and phagocytosis index (mean amount of bacteria phagocytosed per cell) was shown. The conclusion is that, after precise test-adjustment, weak phagocytosis is better detected by phagocytosis percentage and high phagocytosis activity is best shown by phagocytosis index.