Variations of sea surface height(SSH) in the Kuroshio south of Japan are addressed by analyzing 19-year(1993–2011) altimetry data from AVISO. Regionally averaged time series of observed SSH had a rising linear trend at 2.64±0.72 mm/a in this period. By analyzing the power spectra, several periods were recognized in temporal SSH variations, including those around 90 and 360 days. The seasonal cycle of SSH was minimum in winter(February) and maximum in summer(August), with peak-to-peak amplitude about 20.0 cm. The spatial distribution of linear trends was inhomogeneous, with a rising linear trend along the coastline and a tripole structure offshore. Spatial distributions of standard deviation of seasonal SSH show very dynamic activities in the southeast of Kyushu and south of Honshu. Seasonal variations of observed SSH are partially explained by surface buoyancy forcing, local wind forcing and the steric component related to subsurface water beneath the mixed layer. Results show different spatial distributions of correlation coefficient and estimation skill between seasonally observed and modeled SSH, which are calculated from surface buoyancy fl ux, local wind forcing and the steric component related to subsurface water. Of those three, the surface buoyancy fl ux has a greater contribution to variations of observed SSH on the seasonal time scale south of Japan. Variations of sea surface height (SSH) in the Kuroshio south of Japan are addressed by analyzing 19-year (1993-2011) altimetry data from AVISO. Regionally averaged time series of observed SSH had a rising linear trend at 2.64 ± 0.72 mm / a The seasonal cycle of SSH was minimum in winter (February) and maximum in summer (August), with peak- to-peak amplitude about 20.0 cm. The spatial distribution of linear trends was inhomogeneous, with a rising linear trend along the coastline and a tripole structure offshore. Spatial distributions of standard deviation of seasonal SSH show very dynamic activities in the southeast of Kyushu and south of Honshu. Seasonal variations of observed SSH are partially explained by surface buoyancy forcing, local wind forcing and the steric component related to subsurface water beneath the mixed layer. Results show diffe rent spatial distributions of correlation coefficient and estimation skill between seasonally observed and modeled SSH, which are calculated from surface buoyancy fl ux, local wind forcing and the steric component related to subsurface water. of those three, the surface buoyancy fl ux has a greater contribution to variations of observed SSH on the seasonal time scale south of Japan.