The interpretation of single lunar highlands igneous samples is hindered not only by their brecciation but also by a lack of similar intensive work on terrestrial plutonic rocks. Terrestrial studies of cumulate rocks are more extensive and have a well-developed stratigraphic context. In an attempt to establish how much can be learned about the parent magma from single plutonic hand samples, single samples from three successive homogeneous but contrasting layers (anorthosite, anorthositic norite, norite) in the Stillwater Complex Norite 1 subzone were subjected to detailed petrographic and major and trace element chemical analyses. The chemical data show that although the calculated parent magma for each layer is similar, a model of liquidus phaes plus trapped magma model is not simultaneously quite consistent with petrography and chemistry. Further, calculated Eu anomalies and the Fe content of the plagioclases provide evidence that the oxidation of the magma was different for each layer. Such oxidation changes can explain qualitatively the Mg/Fe variations otherwise attributed to trapped liquid. However, it would appear that a reasonable estimate of the parent composition can be made even for single hand samples, such as lunar plutonic highlands rocks. There are uncertainties, and a better understanding of crystallization mechanisms for basic plutonic rocks is needed. Without such understanding lunar samples can never be fully interpreted. The calculated parent for the Stillwater Complex Norite 1 subzone rocks considered is chemically similar to some primitive continental tholeiites, except for its high normative orthopyroxene and negative Eu anomaly, both of which are major differences. Published data show that the Eu anomaly existed in parental magmas even for the lower ultramafic zone, and that most REE data are consistent with common or closely related parents for all rocks of the Stillwater. At least, radically different magmas for different parts of the intrusion do not seem likely.