Interfaces play critical roles in materials and are usually both structurally and compositionally complex microstructural features. The precise characterization of their nature in three-dimensions at the atomic scale is one of the grand challenges for microscopy and microanalysis, as this information is crucial to establish structure–property relationships. Atom probe tomography is well suited to analyzing the chemistry of interfaces at the nanoscale. However, optimizing such microanalysis of interfaces requires great care in the implementation across all aspects of the technique from specimen preparation to data analysis and ultimately the interpretation of this information. This article provides critical perspectives on key aspects pertaining to spatial resolution limits and the issues with the compositional analysis that can limit the quantification of interface measurements. Here, we use the example of grain boundaries in steels; however, the results are applicable for the characterization of grain boundaries and transformation interfaces in a very wide range of industrially relevant engineering materials.