We investigate self-demagnetization effects on magnetic data and develop a comparison of two existing inversion methods as they apply to quantitative interpretation of such data in highly magnetic environments. We begin by evaluating the effect on magnetization direction when susceptibility is a scalar and increases from low values into the realm of self-demagnetization. We show through numerical experiments that susceptibility values of greater than 0.1 SI lead to significant self-demagnetization effects. Second, we show that conventional inversion can perform well for interpretation of self-demagnetization problems with simple source geometries. However, as the geometry becomes more complex in realistically complex problems, this approach can produce poor results and a more robust technique is required. Our numerical experiments indicate that directly inverting amplitude data, which can be derived from total-field magnetic anomaly data and are weakly dependent on magnetization direction, produces superior results when interpreting data from areas with complex geology and high magnetic susceptibilities. We conclude by evaluating the application of our preferred approach on a large field data set exhibiting strong self-demagnetization, multiple source bodies, and complex structures.