In this paper, we analyze magnetic flux leakage (MFL) sensing for the nondestructive evaluation (NDE) of ferromagnetic specimens. Understanding the processes involved in the creation of magnetic flux leakage fields and their measurement is critical to robotic inspection applications. In particular, robotic inspection of energy pipelines uses mobile robots to magnetize sections of the pipe and to measure the MFL signal to detect defects. We study current practices and motivate the need for improvements. To facilitate the analysis, we develop an analytical model to represent the 3-dimensional magnetic flux leakage field due to a surface-breaking defect in the specimen. The model is derived from first principles using the concept of dipole magnetic charge, and uses surface integrals to represent the MFL field as measured by a Hall-effect sensor. Simulations are performed which generate novel results, apart from reproducing experimental results from the literature. The mathematical tractability of the model is exploited to analyze its properties, such as scale–invariance, influence of lift-off, and the tangential MFL component. These properties give new insight into MFL sensing, interpretation, and defect characterization.