Forty melanocytic lesions were examined in a pilot study, which was fo
llowed by a final series of 100 consecutive melanocytic lesions, in or
der to evaluate the inter-observer reproducibility of the histological
criteria proposed for the dysplastic naevus. The specimens were exami
ned in a blind fashion by four observers. Analysis by kappa statistics
showed poor reproducibility of nuclear features, while reproducibilit
y of architectural features was acceptable, improving in the final ser
ies. Consequently, we cannot apply the combined criteria of cytologica
l and architectural features with any confidence in the diagnosis of d
ysplastic naevus, and, until further studies have documented that arch
itectural criteria alone will suffice in the diagnosis of dysplastic n
aevus, we, as pathologists, shall avoid this term.