Systematic development of future scenarios involves identifying key fa
ctors and their different possible values, then selecting combinations
of factor values on which to build scenarios. This article recapitula
tes basic methods of reducing the number of possible combinations of f
actor values by removing futures containing incompatible factor value
pairs, then uses probability and cross-impact measures to further elim
inate improbable futures. Working with the remaining plausible futures
, the aim is to select a small number of them for scenario development
in such a way that there is a balance between the factor values explo
red. The motivation is to explore possible futures while avoiding unin
tentional bias by over-representing some factor values and neglecting
others. The tool used is a mathematical formulation that is easily sol
ved as an integer linear program. The process is illustrated with a sm
all example. (C) 1997 Elsevier Science Inc.