As Virtual Reality (VR) devices become more accessible, a multitude of VR applications engage users in highly immersive virtual environments that feature realistic graphics, real-life scenarios, and self-avatars. Many of these simulations require users to make spontaneous affordance judgments such as stepping over obstacles, passing through gaps, etc. which are shown to be affected by the nature of our self-representation in the virtual world. As the technology for creating self-avatars becomes more widely available, it is important to explore how various affordance judgments are affected by the presence of self-avatars. In this work, we investigate the effects of body-scaled self-avatars on the affordance of passability in a natural setting. We implemented a gender-matched body-scaled self-avatar using HTC Vive trackers and evaluated how passability judgments for a sliding doorway in VR, with and without an avatar, compared to the real world judgments. The results suggest that passability judgments are more conservative in VR as compared to the real world. However, the presence of a self-avatar does not significantly affect passability judgments made in VR. This does not align with previous findings which show that having a self-avatar improves judgments and estimates.