In testing leptospirosis by microagglutination, reproducibility of the results is affected by population age and germ concentration. These undersirable factors can be avoided if antibody identification is made with cultures incubated for 7 to 21 days and adjusted to 50 X 10(6) leptospires per millilitre. Most suitable for the purpose is an analogue mode based on nephelometry and the probability theory. The obtained statistical data furnish calculated and tabulated apparatus values which are used for standardizing with the aid of a newly developed nepheloflask with compression-spring stirrer.