Purpose: Dose perturbation from collimator scattering is not modeled by current treatment planning systems and is usually ignored in proton radiotherapy. In this work we study the dosimetric impact of collimator scattering and possible solutions. Methods: Three sets of apertures with non‐diverging edge, diverging edge and stepped edges that averaged divergence for each single piece, were milled by CURA to produce a 10cm×10cm field at isocenter. Cross‐plane profiles were measured at a shallow (2cm) and deep (7.5cm) depths with EBT3 films, delivered with the Mevion S250 system, which has a relatively shorter SAD (∼2m). Films were irradiated with a fully modulated proton beam with range of 15cm in water. In‐plane profiles were also acquired on a coronal plane along the central axis. Results: Although negligible changes in lateral penumbra were observed in all cases, collimator scattering added up to 11% extra dose near the field edge for the non‐diverging edge aperture. For diverging and stepped edge apertures, extra dose from collimator scattering were reduced to less than 5% at 2cm depth, and was negligible at 7.5cm depth. Deterioration of distal penumbra was observed in both non‐diverging and stepped edged apertures, broadened to 13mm from 8mm, which was measured for the diverging edge aperture. Conclusion: Collimator scattering from apertures not only introduce extra dose near the field edge at shallow depths, but also deteriorates the distal penumbra. A diverging‐cut aperture can eliminate both contaminations. Divergent edges should be considered in apertures when sensitive organs are lateral at shallow depths, or distal to the target. blown MC. The method can be integrated in existing TPS to enable PET‐based treatment verification in the daily clinical routine of proton therapy.