In　this　study,　a　novel　multi-stage　fuzzy　stochastic　programming　(MSFSP)　model　was　developed　for　regional　energy　system　structure　optimization　and　planning　with　energy-water　nexus　under　multiple　uncertainties.　By　cooperating　the　multistage　stochastic　programming　and　fuzzy　theory,　the　developed　model　can　deal　with　energy　system　planning　problems　under　a　mixture　of　probabilistic　and　possibilistic　uncertainties.　Fuzzy　random　scenarios　are　designed　to　express　the　uncertain　future　energy　demand　levels　and　vague　decision　maker＇s　risk　attitude.　A　MSFSP　-based　energy-water　nexus　system　management　model　was　applied　to　a　practical　energy　system　planning　problem　in　an　energy-intensive　and　water-stressed　area,　Tianjin,　China.　Results　of　optimal　capacity　expansion,　power　generation,　and　imported　electricity　strategies　were　obtained;　meanwhile　the　water　resource　availability　effects　and　decision　makers＇　risk　attitude　were　analyzed.　It　was　found　that　water　resource　availability　would　be　a　significant　factor　in　promoting　local　power　structure　and　electricity　generation　in　the　future.　More　serious　water　resource　deficiency　would　lead　to　reduce　local　coal-fired　power　capacity　investment,　while　stimulate　renewable　energy　development　and　increase　imported　electricity　requirement.　All　above　can　facilitate　decision　supports　for　regional　energy　system　planning　from　a　comprehensive　energy-water　nexus　perspective　with　more　sustainable　and　risk-aversion　manners.　(C)　2019　Elsevier　Ltd.　All　rights　reserved.