The visual span during object search in real-world scenes was investigated using the ‘moving window’ experimental paradigm. A repeated measures design was used in which window size was manipulated to create the 7 levels of the independent variable (one control and six experimental), while reaction time formed the main dependent variable. Eye movement data was also recorded to complement the main analysis. Visual span was defined as the window size at which reaction times became indistinguishable from those of the control (no window) condition. Based on previous research findings it was expected that search time would decrease as a function of increasing window size, leading to the determination of the visual span at a window size of 5.4°. While this downward trend was recovered from the reaction time data, which showed a significant effect of window size (p = .00), even the largest window (6.1°) failed to yield reaction times equal to those of the control condition. The likelihood of participants finding the target object did reach control levels in the 5.4° window trials. Analysis of eye movement data indicates that reduced saccade amplitudes in the experimental condition were responsible for the inflated reaction times found.