Target-to-distractor ratio strongly influences performance on typical random array letter cancellation tasks, suggesting that a 'controlled' processing approach is used. This study was designed to determine whether 'automatic' processing could be also demonstrated in the random array cancellation paradigm by changing the perceptual characteristics of the stimuli. Thirty-two healthy subjects sequentially performed four random array cancellation tasks with 50 and 100 stimuli. The letters 'I' or 'O' were targets and 'L' served as the distractor. Performance was measured by the number of correctly canceled targets divided by the time to completion, corrected for accuracy. There was a strong effect of the number of stimuli on forms using I targets (p < .00001), but not for O's (p = .15) Performance scores were lower for I target forms than for O targets. These findings demonstrate that performance approximating 'automatic' processing can also be elicited on clinically useful, office-based, or bedside tests such as random array cancellation.