Computational docking methods have come of age and play a key role in "virtual screening " of large collection of molecules to identify "drug leads" for a given disease target. Exhaustive search of the "binding pocket" of the biological target (protein) for the best fit of the drug molecule is accomplished by a variety of robust schemes that include: Monte Carlo, Genetic Algorithm-based, or Evolutionary programming. The bottlenecks, caveats, ecstasy and agony along the way will be discussed with specific examples.