Both the sudden appearance of an object and sudden changes in existing obje
ct features influence priority in visual search. However, direct comparison
s of these influences have not been made under controlled conditions. In 5
visual search experiments, new object onsets were compared directly with ch
anges in the luminance of old objects. Factors included the luminance contr
ast of items against the background, the magnitude of luminance change, and
the probability that these changes were associated with the target item. N
ew objects were consistently more effective in guiding search, such that a
new item with very low luminance contrast was equivalent to an old item und
ergoing a large change in luminance. An important exception was an old item
changing in contrast and polarity, which was as effective as the appearanc
e of a new object. This indicates that search priority is biased toward obj
ect rather than situational changes.