Well-log-curve shapes and amplitude trends are routinely used to correlate and map formations and reservoirs across petroleum basins or fields. However, the pattern-matching methods typically employed for correlation fail to make full use of the vertical resolution of well-log curves. A new technique, log-curve amplitude slicing, facilitates correlation by automating visualization of spatial trends in log-curve amplitudes. This new technique can be used with any type of log data that is sampled at a regular interval between two distinct correlated surfaces. The method generates a series of subhorizontal slices through the log curves that can be gridded and contoured to show trends and patterns in log-curve amplitudes in map view. The slices represent approximate time lines and are relative chronostratigraphic surfaces. When appropriate logs are used, the slices show the inferred distribution of lithofacies at the time of deposition. The evolution of an area can be investigated by stepping through a series of slices. Application of the log-curve amplitude-slicing technique facilitates correlation because it highlights trends that are not apparent using traditional methods to display and compare log curves. As an example of the log-curve amplitude-slicing technique, gamma-ray log data are used in this article to analyze the evolution of the Michigan basin (United States) during deposition of the Middle Devonian Traverse Group. The Traverse Group consists mostly of carbonates, but the location, relative timing, and extent of significant fine-grained clastic influx to the basin is readily identifiable when the basin's history is reconstructed using the log-curve amplitude-slicing technique.