To study the rate and mechanism of granitic rock weathering, field weathering experiments using granodiorite tablets (diameter 3.5 cm, height 1.1 cm, weight 30 g) were conducted at a catchment over 10 yr. The tablets were exposed at three positions having different weathering conditions: ground surface, above aquifer, and in aquifer. The weight of the tablets decreased linearly with time: 0.022 wt% yr−1 for the ground surface, 0.013 wt% yr−1 above the aquifer, and 0.42 wt% yr−1 in the aquifer. Only minor changes of the tablet surface were observed after 10 yr of weathering at the ground surface and above the aquifer. In contrast, the tablets in the aquifer were remarkably weathered, and weathering of biotite was particularly noticeable, whereas that of plagioclase, K-feldspar, quartz, and hornblende was minor or moderate. The weathering of granodiorite is inferred to proceed by initial dissolution of mineral grain boundary (chemical process) and subsequent detachment of the mineral grain (physical process). To evaluate the amount of weathering caused only by the chemical process, a laboratory dissolution experiment was conducted on the granodiorite tablet using a flow-through reactor at 20 °C and pH 6–7. The obtained rates, 0.0078–0.010 wt% yr−1, were ∼50 times slower than the field rate in the aquifer. This result shows that the contribution of the physical process to granodiorite weathering in the aquifer is very large compared to that of the chemical process, and the relative intensities of the two processes during the 10 yr weathering are estimated to be chemical:physical = 1:∼50.