Molar-tooth structure (MT) is an unusual Precambrian carbonate fabric, which is characterized by variously shaped voids filled with uniform, equant microspar. Their worldwide distribution, in Meso- and early Neoproterozoic rocks may be related to secular changes in substrate rheology and ocean chemistry throughout the Precambrian. Unfortunately, the use of MT as a tool for understanding Proterozoic environments is limited, because its genesis is still poorly understood. Recently, however, Furniss et al. (1998) showed experimentally that gas generated by decaying organic matter within unconsolidated mud, can reproduce MT crack morphologies. This gas expansion hypothesis allows MT morphologies to be modeled as an interaction between the expansive force of the gas and the cohesion of the surrounding substrate. Three scenarios arise: 1) if the expansive force < substrate cohesion, a blob morphology should result; 2) if the expansive force > substrate cohesion, a ribbon morphology should result; and 3) if expansive force >>> substrate cohesion, gas should diffuse through the sediment pore space leaving no apparent void. Each scenario is observed petrographically in MT from the 1.4 Ga Helena Formation, Montana. The greatest variation in MT morphology occurs when cracks interact with sediment of differing cohesiveness. For example, MT cracks commonly track and/or terminate within coarse-grained laminae. When MT cracks terminate within coarse-grained layers (quartz sand, ooids, intraclasts) MT cement commonly fills pore space, forming the primary matrix component. In rare cases, coarse-grained material is observed to have collapsed into underlying cracks prior to precipitation of void-filling MT cement. These petrographic relationships are significant in that they: 1) support MT genesis by gas expansion, 2) suggest that substrate cohesiveness determines crack behavior, and 3) indicate that MT cements formed penecontemporaneously with crack formation, from fluids in contact with the overlying water column.