Two experiments demonstrate that grouping can be strongly influenced by the
presence of figures defined by illusory contours. Rectangular arrays were
constructed in which a central column of figures could group either with th
ose on one side, on the basis of perception of figures defined by illusory
contours, or with those on the other side, on the basis of physically prese
nt inducing elements. In all displays, subjects grouped according to the il
lusory figures significantly more often than for control displays that cont
ained the same inducing elements, but rearranged so that illusory contours
were degraded or eliminated. A second experiment showed that in objectively
defined grouping tasks, subjects grouped faster by illusory figures than b
y inducing elements. These results indicate that grouping can occur after i
llusory contours have been perceived.