In virtual environments, we often interact with avatars that may provide spatial reference frames. Previous studies have shown that such reference frames change the spatial encoding of objects. For example, the same object may be seen as left or right relative to an avatar and this spatial code may change based on the avatar’s perspective. The position of objects from the avatar’s perspective can lead to spatial compatibility effects, indicating that objects are spontaneously coded relative the avatar’s perspective rather than from a person’s own point of view. The present study investigated the role of attention for these avatar-based compatibility effects. Participants were asked to respond to the color of vertically presented stimuli in a Simon task. The stimuli were presented next to an avatar that provided a reference frame to code the position of the stimuli as either left or right. In half of trials, the avatar flashed red at the time of stimulus presentation to draw the participants’ attention to it (pop condition). In the other half, the avatar remained unchanged (control condition). The results showed a Simon effect from the avatar’s perspective in both conditions. Importantly, this Simon effect was significantly larger in the pop condition, indicating that attentional shifts towards the avatar increased the impact of the reference frame provided by the avatar.