The last few years have seen enormous damage and lossof life from climate and weather phenomena. The mostdamaging events have included the severe 1997/98 ElNiño (with its near-global impacts), HurricaneMitch, and floods in China in mid-1998. What have welearnt regarding the causes, variability, andpredictability, of these phenomena? Can we predict theoccurrence of these extreme events, and therebymitigate their damage? This paper reviews what we havelearnt in the last decade or so regarding thepredictability of these climate and weather extremes.The view starts with the largest (El Niño) scales,and works towards the scale of individualthunderstorms. It focuses on the practical outcomes ofour improved knowledge with regard to decreasing theimpact of natural disasters, rather than describing indetail the scientific knowledge underlying theseoutcomes. The paper concludes with a discussion ofsome of the factors that still restrict our ability tomitigate the deleterious effects of atmospheric andclimatic hazards.