Two experiments were directed at distinguishing associative and simila
rity-based accounts of systematic differences in categorization time f
or different items in natural categories. Experiment I investigated th
e correlation of categorization time with three measures of instance c
entrality in a category. Production frequency (PF), rated typicality,
and familiarity from category norms for British participants (Hampton
& Gardiner, 1983) were used to predict mean categorization times for 5
31 words in 12 semantic categories. PF and typicality (but not familia
rity) were found to make significant and independent contributions to
categorization time. Error rates were related only to typicality (apar
t from errors made to ambiguous or unknown items). Experiment 2 provid
ed a further dissociation of PF and typicality. Manipulating the diffi
culty of the task through the relatedness of the false items interacte
d primarily with the effect of typicality on categorization time, wher
eas, under conditions of easy discrimination, prior exposure to the ca
tegory exemplars affected only the contribution of PF to the decision
time. The dissociation of typicality and PF measures is interpreted as
providing evidence that speeded categorization involves both retrieva
l of associations indexed by PF and a similarity-based decision proces
s indexed by typicality.