Abstract A large earthquake (Mw=7.7) along a plate boundary occurred in the south of Java Island on July 17, 2006, and caused a significant tsunami. We made GPS observations and tsunami heights measurements during the period from July 24 to August 1, 2006. The earthquake seems to be due to an interplate low angle reverse faulting, though there might be a possibility of high angle faulting within the subducting lithosphere. Crustal deformation distribution due to the earthquake, aided by tsunami heights measurements, might clarify which would be the case. We occupied 29 sites by GPS in the area of southern Java encompassing the area from 107.8 E to 109.50 E. These sites were occupied once before the earthquake. However, we were not able to detect significant coseismic displacements. The obtained displacements, most of which span several years, show ESE direction in ITRF2000 frame. This represents the direction of Sunda block motion. The tsunami heights measured at 11 sites were 6–7 m along the southern coast of Java and indicate that the observed heights are systematically higher than those estimated from numerical simulations that are based on seismic data analysis. This might suggest that fault offsets might have been larger—nearly double—than those estimated using seismic analysis. These results lead us to an idea that the rupture was very slow. If this is the case, the earthquake might have been a “tsunami earthquake” that is similar to the one that occurred on June 2, 1994 in the east of the present earthquake.