Monsanto employs several pragmatic approaches for evaluating the toxic
ity of mixtures. These approaches are similar to those recommended by
many national and international agencies. When conducting hazard and r
isk assessments, priority is always given to using data collected dire
ctly on the mixture of concern. To provide an example of the first tie
r of evaluation, actual data on acute respiratory irritation studies o
n mixtures were evaluated to determine whether the principle of additi
vity was applicable to the mixture evaluated. If actual data on the mi
xture are unavailable, extrapolation across similar mixtures is consid
ered. Because many formulations are quite similar in composition, the;
toxicity data from one mixture can be extended to a closely related mi
xture in a scientifically justifiable manner. An example of a family o
f products where such extrapolations have been made is presented to ex
emplify this second approach. Lastly, if data on similar mixtures are
unavailable, data on component fractions are used to predict the toxic
ity of the mixture. In this third approach, process knowledge and scie
ntific judgement are used to determine how the known toxicological pro
perties of the individual fractions affect toxicity of the mixture. Th
ree examples of plant effluents where toxicological data on fractions
were used to predict the toxicity of the mixture are discussed. The re
sults of the analysis are used to discuss the predictive value of each
of the above mentioned toxicological approaches for evaluating chemic
al mixtures. Copyright (C) 1997 Published by Elsevier Science Ltd.