The neutron absorption macroscopic cross-section, Sigma , is measured routinely by neutron porosity tools and, although rarely presented as a logging curve in its own right, is used indirectly for the estimation of (neutron) porosity. One of the reasons that this primary measurement is not often employed directly in petrophysical analysis is the difficulty of interpretation. In particular, little is known about the range of Sigma values for common lithologies, or exactly what information the measurement is providing. In this contribution we demonstrate that excellent estimates of Sigma can be calculated provided that the chemistry of a sample is known in sufficient detail. When applied to a range of geochemical reference materials, it becomes apparent that the minor and trace elements present may have a profound effect on the Sigma value of a sample, and, in turn, on the interpretation of neutron porosity measurements. Using this approach we present Sigma data for basaltic and ultrabasic rocks, and model the change in Sigma with alteration. Alteration is considered in these models as an increase in alteration minerals (which are mainly clays, but also carbonates and zeolites in basic rock alteration) and changes in the trace-element chemistry of the rocks. Of the trace elements, boron and some of the rare-earth elements are of particular importance. Modelling the variation in Sigma with these mineralogical and compositional changes indicates that increases in boron are the most important of these factors in increasing Sigma ; this is enhanced by the alteration, particularly to clay phases, which generally accompanies an increase in boron. These models suggest that a Sigma log should be able to act as a proxy for alteration trends in basic and ultrabasic crystalline rocks, and a quantitative model for such alteration is described.