The study design and interpretation criteria for the rat liver UDS ass
ay have evolved over recent years. A database from a single laboratory
is presented to demonstrate the range of responses encountered whilst
screening 71 industrial and agrochemical compounds for potential geno
toxicity, a total of 280 experiments. With a large data set it was pos
sible to delineate clearly the normal responses of the system and appl
y robust criteria for the evaluation of potentially positive compounds
. The rat liver UDS assay proved to be a reproducible, reliable and se
nsitive technique for the evaluation of genotoxicity in vivo., Experim
ents with negative control values of (N-C) greater than zero would hav
e been rejected, however the highest negative control value observed w
as (N-C) = -1.1. Evaluation of data using the current United Kingdom E
nvironmental Mutagen Society (UKEMS) recommended procedures resolved 9
6% of studies by simple inspection. In the remainder of studies data w
ere clarified by additional experiments under modified conditions. It
is unlikely that outcomes would have been improved upon by direct stat
istical analysis which would require an expanded experimental design.
These data confirm the utility of the pragmatic approach of the curren
t UKEMS guidelines.