Functional lung imaging is of great importance for diagnosis and follow-up of prevalent lung diseases. To this end, novel pulse sequences and post-processing techniques have been previously proposed to assess pulmonary functions. However, one overlooked aspect has been the parallel imaging reconstruction. Here, we propose the use of an advanced reconstruction scheme based on Low-Rank Modeling of Local k-Space Neighborhoods (LORAKS). In vivo results are provided to demonstrate the performance of LORAKS compared to commonly used GRAPPA reconstructions. Preliminary results indicate that LORAKS can be a viable option for improving reconstruction quality in pulmonary functional imaging.