We investigated the haptic cuing of visual attention using spatially-predictive (75% valid) and spatially-nonpredictive (25%) haptic cues. The participants performed a visual change detection task immediately following a haptic spatial cue whose location corresponded to one of the four visual quadrants. The participants were explicitly instructed to use the spatially-predictive haptic cues but to try and ignore the spatially- nonpredictive cues. In addition to reaction time (RT) data, we recorded participants' eye-position in order to provide a direct measure of overt visual attention. The results indicated that the spatially-predictive haptic cues reduced the amount of time taken to detect the visual changes, as expected. The spatially-nonpredictive cues increased visual search latencies, indicating that the cues could not be ignored completely. There was also evidence that haptic cuing served to alert the participants resulting in an overall reduction of response latencies.