Abstract: The method for determining endotoxin concentrations in samples and, if present, interfering factors by adding defined amounts of endotoxin to the sample and computational evaluation of the reaction kinetics can be improved by adding, to at least two identical preparations of a sample, different amounts of endotoxin. These endotoxin amounts, however, are at least ten times the amount of the endotoxin to be determined in the sample. By determining a regression line as the reference curve from the measured values, the endotoxin amount of a preparation of the sample to which no endotoxin has been added is determined. The effect caused by interfering factors and the endotoxin neutralization capacity can be determined from the slope and position of the regression line. The absorbance in a turbidimetric measurement may be enhanced by using a small amount of a chromogenic substrate whereby a reduction in the amount of the sample preparation to be assayed will be possible.