The Quantitative Structure Activity Relationships (QSAR) methods were originally developed for the needs of drug and pesticide design.

In more recent years, (Q)SAR (i.e., both qualitative and quantitative) methods have been developed as alternative to the experimental assays for predicting toxicity.

The identification of the Structural Alerts (SA) , i.e., structural motifs known to be causative of the toxicological effects, is of special interest because: a) they are powerful tools for directly identifying potentially genotoxic chemicals; and b) they can be used for grouping together toxicologically similar chemicals. These mechanistically-based Categories of chemicals can be regulated as a whole, or can be used for Read-Across predictions. Read-Across predictions fill data gaps of a chemical of interest, with toxicological data from similar chemicals.

Another important area of predictive QSAR methods includes mathematical, statistically verified models that link the toxicological properties of chemicals with their structural/ physical chemical characteristics.