A complete and optimized scheme of lettered marine isotope substages spanning the last 1.0 million years is proposed. Lettered substages for Marine Isotope Stage (MIS) 5 were explicitly defined by Shackleton (1969), but analogous substages before or after MIS 5 have not been coherently defined. Short-term discrete events in the isotopic record were defined in the 1980s and given decimal-style numbers, rather than letters, but unlike substages they were neither intended nor suited to identify contiguous intervals of time. Substages for time outside MIS 5 have been lettered, or in some cases numbered, piecemeal and with conflicting designations. We therefore propose a system of lettered substages that is complete, without missing substages, and optimized to match previous published usage to the maximum extent possible. Our goal is to provide order and unity to a taxonomy and nomenclature that has developed ad hoc and somewhat chaotically over the decades. Our system is defined relative to the LR04 stack of marine benthic oxygen isotope records, and thus it is grounded in a continuous record responsive largely to changes in ice volume that are inherently global. This system is intended specifically for marine oxygen isotope stages, but it has relevance also for oxygen isotope stages recognized in time-series of non-marine oxygen isotope data, and more generally for climatic stages, which are recognized in time-series of non-isotopic as well as isotopic data. The terms “stage” and “substage” in this context are best considered to represent climatostratigraphic units, and thus “climatic stages” and “climatic substages”, because they are recognized from geochemical and sedimentary responses to climate change that may not have been synchronous at global scale.