The hydrologic effects of forest use and reforestation of degraded lands in the humid tropics has implications for local and regional hydrologic services but such issues have been relatively less studied whencompared to the impacts of forest conversion. In particular, the ‘‘infiltration-evapotranspiration trade-off’’hypothesis which predicts a net gain or loss to baseflow and dry-season flow under both, forest degradation or reforestation depending on conditions has not been tested adequately. In the Western Ghats ofIndia, we examined the hydrologic responses and groundwater recharge and hydrologic services linkedwith three ecosystems, (1) remnant tropical evergreen forest (NF), (2) heavily-used former evergreen forest which now has been converted to tree savanna, known as degraded forest(DF), and (3) exotic Acaciaplantations (AC,Acacia auriculiformis) on degraded former forest land. Instrumented catchments rangingfrom 7 to 23 ha representing these three land-covers (3 NF, 4 AC and 4 DF, in total 11 basins), were established and maintained between 2003 and 2005 at three sites in two geomorphological zones, Coastal andUp-Ghat (Malnaad). Four larger (1–2 km2) catchments downstream of the head-water catchments in theMalnaad with varying proportions of different land-cover and providing irrigation water for areca-nutand paddy rice were also measured for post-monsoon baseflow. Daily hydrological and climate datawas available at all the sites. In addition, 36 min data was available at the Coastal site for 41 days as partof the opening phase of the summer monsoon, June–July 2005.Low potential and actual evapotranspiration rates during the monsoon that are similar across all landcover ensures that the main control on the extent of groundwater recharge during the south-west monsoon is the proportion of rainfall that is converted into quick flow rather than differences in evapotranspiration between the different land cover types. The Flow duration curves demonstrated a higherfrequency and longer duration of low flows under NF when compared to the other more disturbed landcovers in both the Coastal and Malnaad basins. Groundwater recharge estimated using water balanceduring the wet-season in the Coastal basins under NF, AC and DF was estimated to be 50%, 46% and35% respectively and in the Malnaad it was 61%, 55% and 36% respectively. Soil Water Infiltration andMovement (SWIM) based recharge estimates also support the pattern (46% in NF; 39% in AC and 14%in DF). Furey–Gupta filter based estimates associated with the Coastal basins also suggest similar groundwater recharge values and trends across the respective land-covers: 69% in NF, 49% in AC, and 42% in DF.Soil water potential profiles using zero flux plane methods suggest that during the dry-season, naturalforests depend on deep soil moisture and groundwater. Catchments with higher proportion of forestcover upstream were observed to sustain flow longer into the dry-season. These hydrologic responsesprovide some support towards the ‘‘infiltration-evapotranspiration trade-off’’ hypothesis in which differences in infiltration between land-cover rather than evapotranspiration determines the differences ingroundwater recharge, low flows and dry-season flow. Groundwater recharge is the most temporally stable under natural forest, although substantial recharge occurs under all three ecosystems, which helps to sustain dry-season flow downstream in higher order streams that sustain local communities and agroecosystems. In addition to spatial scale effects, greater attention also needs to be given to the role ofhydrogeology within the context of the above hypothesis and its implications for hydrologic services.