Since the inception of the oil boom in North Dakota, the Williston basin has witnessed a tremendous growth in horizontal drilling and completion activity primarily targeting the Bakken and Three Forks formations. Although the activity in the basin is maturing in terms of our understanding rock quality and completion quality, there is a wide variation of these indices within the basin from one field to another. Some of these variations are clearly noticeable in parameters such as thicknesses of the shale barriers, pore pressure gradients, reservoir permeabilities, porosities and stress gradients. The combined impact of these parameters has a huge impact on key decisions including, but not limited to, completion methodologies, types of proppants and fluids used for completion, number of fracturing stages in the lateral, number of perforation clusters per stage, and well spacing. This paper discusses the evolution of stimulation strategies and completion practices in the Williston basin since 2009. Operators have experimented with cemented and uncemented laterals; sliding sleeves and plug-and-perf completions; lateral lengths ranging from 5,000 to 10,000 ft; perforation clusters ranging from one to six per stage; crosslinked, hybrid, and slickwater fluid systems; proppants ranging from sand to ceramic, etc. The consequent impacts of these variations on well completion pressure responses and long-term production have been mixed. As part of the work covered in this paper, the differences between various completion methodologies and their impact on the stimulation strategies have been discussed in a chronological order.Although there is no single optimized design for the entire basin, experimentation of multiple methods and technical interpretation of various fracture and production models have provided us with a strong foundation to narrow down our practices to the most successful and repeatable ones across all the fields in the Bakken and Three Forks formations. The paper also covers how real-field measurements such as diagnostic fracture injection tests (DFITs), microseismic data, radioactive or chemical tracers, bottomhole pressure gauges, and interference experiments combined with log measurements such as magnetic resonance, acoustic logs, and elemental spectroscopy can provide us with a strong base for building and calibrating reservoir models that are reliable and reasonable.The paper covers technical differences between sliding sleeves and plug-and-perf completions; differences between crosslinked, slickwater, and hybrid designs and their impact on fracture geometries; effect of using different proppant types; and ways to optimize the number of fracturing stages and proppant and fluid volumes. As part of the study, the importance of geomechanics in understanding planar versus complex fracture geometries is discussed to close the loop with reservoir simulation models.