Abstract Technological advances in drilling equipment, directional control and mud systems and a better understanding of reservoir properties have resulted in the drilling of an increasing number of highly deviated and horizontal wells. Along with these wells come the increasing borehole instability and sand production potentials under normal in-situ stress conditions (when the stresses are vertical and horizontal, with vertical greater than both horizontal stresses). Conventional guidelines for horizontal well design have been to drill in the direction parallel to the minimum principal stress, which reduces the difference between the principal stresses perpendicular to the borehole, and results in a minimum stress concentration around the borehole. Such guidelines are based on an effort to reduce the required borehole pressure to maintain zero breakout. Research on borehole breakout and borehole stability in the past fifteen years has shown that a stable breakout is characterized by a stress distribution which transfers the high stress concentration from the opening to the zone near the tip of the breakout, and causes a quasi-hydrostatic stress state in that zone. Breakout geometry is therefore more stable than circular geometry. Recent experimental studies on utilizing breakout geometry to enhance tunnel stability have also shown that breakout geometry has a much greater "excavation strength" than circular geometry under the same stress configurations. Breakout geometry provides a "self-supportive" function that allows a smaller degree of failure potential around the borehole than other excavation geometries. Fundamental stability mechanisms for stable breakout are revisited and the feasibility of drilling parallel to the intermediate principal stress direction under stable conditions investigated through numerical simulations. The results show that under most in-situ conditions, drilling in the intermediate principal stress direction and allowing some degree of breakout affords tmproved stability during drilling and subsequent production, provided that the breakout generated failure materials can be removed efficiently. The mud weight required to maintain stability is substantially less for this type of borehole configuration that leads to reduced formation damages, and allows for the generation of a single hydraulic fracture parallel to the borehole rather than multiple fractures perpendicular to the borehole (assuming it is drilled in the direction of the minimum principal stress). The cost savings and possible production enhancement derived from such drilling practices could result in a dramatic economic impact on the petroleum industry. P. 395