Theoryandevidenceindicatethattreesandothervegetationinfluencetheatmosphericwater-cycleinvariousways.Theseinfluencesaremoreimportant,morecomplex,andmorepoorlycharacterisedthaniswidelyrealised.Whilethereislittledoubtthatchangesintreecoverwillimpactthewater-cycle,thewiderconsequencesremaindifficulttopredictastheunderlyingrelationshipsandprocessesremainpoorlycharacterised.Nonetheless,asforestsarevulnerabletohumanactivities,theselinkedaspectsofthewater-cyclearealsoatriskandthepotentialconsequencesoflargescaleforestlossaresevere.Here,fornon-specialistreaders,Ireviewourknowledgeofthelinksbetweenvegetation-coverandclimatewithafocusonforestsandrain(precipitation).Ihighlightadvances,uncertaintiesandresearchopportunities.Therearesignificantshortcomingsinourunderstandingoftheatmospherichydrologicalcycleandofitsrepresentationinclimatemodels.Abetterunderstandingoftheroleofvegetationandtree-coverwilreducesomeoftheseshortcomings.Ioutlineandilustratevariousresearchthemeswheretheseadvancesmaybefound.Thesethemesincludethebiologyofevaporation,aerosolsandatmosphericmotion,aswellastheprocessesthatdeterminemonsoonsanddiurnalprecipitationcycles.Anoveltheory-the‘bioticpump’-suggeststhatevaporationandcondensationcanexertamajorinfluenceoveratmosphericdynamics.Thistheoryexplainshowhighrainfallcanbemaintainedwithinthosecontinentalland-massesthataresufficientlyforested.Feedbackswithinmanyoftheseprocessescanresultinnon-linearbehavioursandthepotentialfordramaticchangesasaresultofforestloss(orgain):forexample,switchingfromawettoadrylocalclimate(orvisa-versa).Muchremainsunknownandmultipleresearchdisciplinesareneededtoaddressthis:forestscientistsandotherbiologistshaveamajorroletoplay.Newideas,methodsanddataofferopportunitiestoimproveunderstanding.Expectsurpr
