Human volunteer studies of skin irritation have been carried out for d
ecades, both for research and for safety evaluation purposes. For the
majority of this time, and consequently for the majority of the studie
s, assessment of the skin reactions has been made visually. Typical en
dpoints include erythema, oedema, dryness and scaling, some or all of
which would be rated on a simple scale, eg 0, +/-, +, ++, +++. Such ap
proaches can be criticized as subjective, of poor reproducibility, lac
king in sensitivity and highly variable between observers and/or insti
tutions. In consequence, instrumental methods of assessment have been
strongly promoted and do indeed offer several advantages, not least th
eir objectivity. However, it is possible to use the human eye, which i
s a very sensitive tool, to make detailed, accurate and reproducible d
escriptions of skin irritation reactions. To achieve this, it is neces
sary to give prolonged and thorough training to each observer. In this
paper, 3 examples of human volunteer studies, in which different pair
s of trained observers independently carried out double blinded scorin
g of the irritation reactions, are reported. The grading patterns prod
uced were almost identical; statistical analysis showed that properly
trained observers are in fact able to reliably measure a grade of eryt
hema to within +/-1 on a 10 point scale; 97.6% of scores were within 2
grade points on this scale. These results provide evidence that visua
l scoring can be sensitive, reliable and reproducible within a testing
institution.