Two experiments examined performance in a sequence learning task. Participa
nts were trained on a repeating sequence which was presented as a visual di
splay and learning was measured via the increase in reaction time to respon
d to a new sequence. Some participants made a response to each stimulus whi
le others merely observed the sequence. In Experiment 1 participants respon
ding to the display via a keypress showed learning, but those merely observ
ing did not. Five possible reasons for the failure to find observational le
arning were considered and the Experiment 2 attempted to resolve these. Thi
s second experiment confirmed the findings of Experiment 1 in a non-spatial
sequence display using a cover story which encouraged attention to the dis
play but not rule-search strategies. The results are discussed in relation
to applied and theoretical aspects of implicit learning.