Increased atmospheric CO2 concentration is widely being considered as the main driving factor that causes the phenomenon of global warming, due to the ever-boosting use of fossil fuels. In this study, a fuzzy-stochastic programming model with soft constraints (FSP-SC) is developed for electricity generation planning and greenhouse gas (GHG) abatement in an environment with imprecise and probabilistic information. The developed FSP-SC is applied to a case study of long-term planning of a regional electricity generation system, where integer programming technique is employed to facilitate dynamic analysis for capacity expansion within a multi-period context to satisfy increasing electricity demand. The results indicate different relaxation levels can lead to changed electricity generation options, capacity expansion schemes, system costs, and GHG emissions. Several sensitivity analyses are also conducted to demonstrate that relaxation of different constraints have different effects on system cost and GHG emission. Tradeoffs among system costs, resource availabilities, GHG emissions, and electricity-shortage risks can also be tackled with the relaxation levels for the objective and constraints. Copyright (c) 2012 John Wiley & Sons, Ltd.