Visual working memory (VWM) is limited to approximately four items. Space-based, object-based, and feature-based attention have all been shown to enhance VWM. However, to our knowledge, no studies have examined them within a single task. In three experiments, we examined the effects of these three types of attention using a change detection task. In each experiment, participants saw a spatial cue, followed by a memory display consisting of four coloured circles. After a retention interval, a memory probe would appear. The task was to determine whether the probe matched the colour of a stimulus in the memory display. In Experiment 1, the memory stimuli were four different coloured circles and were grouped by task-irrelevant background rectangles. The probe stimulus could appear in the cued location, in the other end of the cued rectangle, or in a different rectangle of equal distance. Accuracy was higher in the cued location compared to both the same-rectangle location and the different-rectangle location, indicating a spatial attention effect. Accuracy was also higher in the same-rectangle location than in the different-rectangle location, suggesting an object effect. In Experiment 2, we additionally grouped two of the memory stimuli by colour, and the cue always indicated one of the colour grouped stimuli. The results showed that the stimuli with the grouped colour were responded to more accurately than the stimuli with ungrouped colours. Furthermore, for the stimuli sharing the same colour, there was no difference in performance regardless of whether the probed stimulus was cued or not, or within the same object as the cue or in a different object from the cue. These findings indicate a perceptual grouping by colour effect with no evidence of spatial or object-based effects. Experiment 3 further examined whether the spatial attention effect, or its absence in Experiment 2, was masked by the colour grouping effect due to the cue always appeared at the location of a grouped stimulus. We tested this hypothesis by manipulating the relationship between the location of the cue and that of a colour grouped stimulus so that the cue indicated the location of a grouped stimulus in some trials (cue-grouped), and the location of an ungrouped stimulus in other trials (cue-ungrouped). Among other results, the most important finding was a colour grouping effect with no spatial effect in the cue-grouped condition, but a colour grouping effect and a spatial attention effect in the cue-ungrouped condition. This pattern of data suggests that the absence of the spatial effect in the present study was due to spatial attention capturing both the cued stimulus and the same-coloured stimuli when the cue appeared at the location of a grouped stimulus. Taken together, these experiments clarify the conditions under which different attentional effects on VWM manifest, and how space-based, object-based and feature-based attentional effects work together within the same task to enhance VWM performance.