Introduction: Magnetic Particle Imaging (MPI) is an emerging medical imaging modality that detects super-paramagnetic particles exploiting their nonlinear magnetization response. Spatial encoding can be realized using a Field Free Line (FFL), which is generated, rotated and translated through the Field of View (FOV) using a combination of magnetic gradient fields and homogeneous excitation fields. When scaling up systems and/or enlarging the FOV in comparison to the scanner bore, ensuring homogeneity and linearity of the magnetic fields becomes challenging. The present contribution describes the first comprehensive, systematic study on the influence of magnetic field imperfections in FFL MPI. Methods: In a simulation study, 14 different FFL scanner setups have been examined. Starting from an ideal scanner using perfect magnetic fields, defined imperfections have been introduced in a range of configurations (nonlinear gradient fields, inhomogeneous excitation fields, or inhomogeneous receive fields, or a combination thereof). In the first part of the study, the voltage induced in the receive channels parallel and perpendicular to the FFL translation have been studied for discrete FFL angles. In the second part, an imaging process has been simulated comparing different image reconstruction approaches. Results: The induced voltage signals demonstrate illustratively the effect of the magnetic field imperfections. In images reconstructed using a Radon-based approach, the magnetic field imperfections lead to pronounced artifacts, especially if a deconvolution using the point spread function is performed. In images reconstructed using a system function based approach, variations in local image quality become visible. Conclusion: For Radon-based image reconstruction in FFL MPI in the presence of inhomogeneous and nonlinear magnetic fields, artifact correction methods will have to be developed. In this regard, a first approach has recently been presented by another group. Further research is required to elucidate the influence of magnetic field imperfections in MPI using a system function based approach.