The objective was to provide a more comprehensive understanding of how people make numerical comparative judgments when digits are contained in numbers with redundant leading or following zeros, e.g., 00080 and 800.000. These sequences of numbers often appear on computer display terminals (VDT) as line numbers, but surprisingly little research has been done on this. The experiment manipulated three aspects of numerical stimuli: (1) redundant leading zeros, (2) redundant following zeros, and (3) length of string of digits. The subjects were to push one of the two button-switches to respond whether two stimulus numbers shown on the computer screen were equal or unequal. The target stimuli contained several forms of redundant zeros, and each performance was assessed by response RT of the subjects. Analysis indicated five significant findings: (1) Redundant leading zeros hindered performance, (2) The effect of redundant following zeros depended on the stimulus type, (3) Over-all, longer digits took more processing time, (4) The RTs for the second-block trials were significantly faster than the first, and (5) Task performance was facilitated when the redundant zero representations were identical in both stimuli of a pair. Nonlexicographic processing seems to occur when feature identification can be used for numerical identification, that is, when the format is consistent. The research has implications for those in computer science to provide numerical formats which make comparative judgments as easy as possible.