This paper analyzes grandfather rules (GR) within the context of the 'optimal tax reform' problem and a reform that eliminates differential capital taxation. Two general conclusions are suggested. First, GR can be a very effective reform implementation tool. Appropriately structured GR can convert potential 'losers' from reform into gainers so that the reform becomes Pareto-improving; alternatively, GR can often be used to distribute the pins from reform more equitably. Second, GR should nevertheless be used with caution. Market adjustments often imply that losses are sufficiently small that little grandfathering is needed; moreover, the effects of GR are very sensitive to the length of time they are in force, and their implementation may lead to efficiency losses.