The objective of this paper is to evaluate the clinical impact of biological uncertainties in small field proton therapy due to the assumption of using a constant relative biological effectiveness (RBE) value of 1.1 (RBE-fixed) compared to a variable RBE (RBE-weighted). In this context the impact of the applied range margin was investigated. Eight patients with arteriovenous malformation (AVM) treated with proton radiosurgery were selected due to the small target volume. Dose distributions were compared for RBE-weighted and RBE-fixed. The impact of RBE was assessed using Monte Carlo (MC) dose calculations for stereotactic doses and doses of 2 Gy(RBE). Four different α/β ratios were investigated. Additionally, dose distributions were recalculated with reduced range margins. Applying variable RBE values for stereotactic doses resulted in an increase in the mean dose of 1.6% for a low α/β of 2 Gy, but a decrease of 2.6% for an α/β of 10 Gy. However, the mean dose increased to 17.1% and 2.1% for doses of 2 Gy(RBE) and α/β of 2 Gy and 10 Gy, respectively. Reducing range margins from 3.5% + 1 mm to 2.5% + 1 mm resulted in negligible difference in the mean RBE within the target, or 0.1% for stereotactic doses and 0.3% for doses of 2 Gy(RBE). Larger differences were seen for a range reduction to 0% + 1 mm, i.e. 1.1% and 3.0% for stereotactic doses and doses of 2 Gy(RBE), respectively. Because potential RBE effects are typically more pronounced in the distal part of a field, a bigger clinical impact of RBE uncertainties in small fields is expected. Our study shows that this could be significant for tissues with low α/β and a small dose per fraction. The uncertainty in RBE due to the uncertainty associated with the α/β ratio seems larger than the impact of the applied range uncertainty margin on RBE.