Understanding　the　seasonal　behaviour　of　a　subtropical　forest　and　its　inter-annual　variation　is　crucial　to　understanding　and　monitoring　its　ecosystem　function　in　the　context　of　global　warming.　Based　on　the　Moderate　Resolution　Imaging　Spectroradiometer　(MODIS)　Enhanced　Vegetation　Index　dataset,　a　wavelet　transform　method　was　used　to　investigate　the　inter-annual　variations　of　vegetation　phenology　in　a　subtropical　mountain　and　hill　region　in　Fujian,　China,　during　2001-2010.　The　results　show　a　distinct　inter-annual　variation　of　vegetation　phenology　related　to　climate　variability　even　if　most　areas　presented　non-significant　trends.　The　start　dates　significantly　advanced　and　end　dates　delayed　in　2003　and　2008,　due　to　anomalously　warm　conditions.　There　was　generally　a　gradient　of　increasing　start　dates,　and　earlier　end　dates　of　vegetation　growing　season,　due　to　colder　temperatures　at　higher　altitudes.　However,　the　altitudinal　phenology　relationship　also　depends　on　its　corresponding　rainfall　conditions.　Earlier　start　dates　were　observed　at　higher　altitudes　during　rainfall　deficit　years　such　as　2008,　which　coincides　with　relatively　abundant　rainfall　at　higher　altitudes.　This　paper　reveals　that　vegetation　phenology　was　coupled　with　altitudinal　gradient,　with　distinct　responses　at　different　combinations　of　alternate　temperature　and　precipitation　conditions　variability.　©　2013　The　Author(s).