In many instances of risk assessment, one has to estimate the potentia
l risk of chemicals using limited experimental data, or no empirical d
ata at all. In such cases, the use of non-empirical parameters, which
can be calculated directly from structure, is a viable option for the
risk assessor, Graph invariants have been used in predicting propertie
s of congeneric sets of chemicals and determining structural similarit
y/dissimilarity of molecules. In this paper we have used (a) topologic
al parameters in predicting mutagenicity of a diverse set of 520 chemi
cals and (b) graph theoretic parameters in quantifying structural simi
larity for a selection of analogs. The results of these analyses are p
resented along with a critical discussion of the utility and limitatio
ns of these methods.