The current study assessed the performance of a patient with right neglect (M.A.H.) across various manipulations of the flanker paradigm. When required to identify a central target in the presence of a unilateral flanker, M.A.H. responded to the flanking distractor on the left (ipsilesional) side as if it were the target, even when the flanker appeared at left peripheral locations. A right (contralesional) flanker did not affect central identification performance (Experiment 1). The "ipsilesional capture" effect persisted when pretrial location markers were introduced to make the flanker and target locations more clearly defined (Experiment 2). However, when the ipsilesional flanker appeared simultaneously with a contralesional flanker, central target detection improved to ceiling (Experiment 3). Interestingly, with these three-stimulus displays, congruency effects in reaction time only occurred in relation to the flanker on the contralesional side (Experiment 3), suggesting impaired response selectivity to ipsilesional stimuli. Congruency effects were produced on both sides only when the two flanking distractors grouped together (by both onset and offset, Experiment 4) and when the ipsilesional flanking distractor grouped with the target by onset (lone contra offset, Experiment 4). The results are attributed to ipsilesional capture in central target detection, which is offset by temporal grouping processes when another stimulus appears on the contralateral side.