Sequence stratigraphy and sequence boundary interpretations for Mississippian Madison stratigraphic units in the greater Williston basin area illustrate some of the challenges for sequence stratigraphy in cratonic carbonate-evaporite depositional systems. Some of the most significant interpretation differences occur in middle Madison strata that correlate with the upper Tournaisian Frobisher-Alida interval in North Dakota which, in the present paper, is considered to represent a third-order (2-3 million-year) sequence. Widely published surface studies in Wyoming place a third-order sequence boundary near the base of a regional solution breccia bed that corresponds with the Frobisher-Alida evaporite. In the subsurface, it can be demonstrated that most of the evaporite formed during regressive deposition and a stratal surface near the base of the evaporite is not a viable choice for a third-order sequence boundary. The Frobisher-Alida third-order maximum regressive surface separates underlying regressive deposits from overlying transgressive deposits and it broadly corresponds with conventional sequence boundary placement. This is a subtle surface that may lie either within or on top of the Frobisher-Alida evaporite or solution breccia equivalent. The third-order maximum flooding surface is placed on a widely mappable submarine unconformity that occurs as a sharp contact on an eroded Thalassinoides-burrowed firmground or a breccia-conglomerate lag deposit. This surface separates transgressive deposits from overlying open-marine strata. Fourth-order Frobisher-Alida sequences are defined in landward areas by the cyclic alternation of anhydrite-dominated strata, with dolomitic and siliciclastic marker beds. Both low-magnitude sea-level movements and the paleoclimatic change from a hot, arid climate (anhydritic strata) to a hot, subhumid climate (dolomitic and siliciclastic marker beds) are responsible for the cyclicity. Abrupt sea-level falls sometimes caused exposure that generated local but prominent fourth and fifth-order subaerial erosion surfaces. In this setting, caution should be employed if subaerial erosion is used to indicate the location of a third-order sequence boundary.