Understanding low impact development (LID) planning and best management practices (BMPs) effects on recharge is important because of the increasing use of LID BMPs to reduce storm water runoff and improve surface-water quality. LID BMPs are microscale, decentralized management techniques such as vegetated systems, pervious pavement, and infiltration trenches to capture, reduce, filter, and slow storm water runoff. Some BMPs may enhance recharge, which has often been considered a secondary management benefit. Here we report results of a field and HYDRUS-2D modeling study in San Francisco, California, USA to quantify urban recharge rates, volumes, and efficiency beneath a LID BMP infiltration trench and irrigated lawn considering historical El Niño/Southern Oscillation (ENSO) variability and future climate change using simulated precipitation from the Geophysical Fluid Dynamic Laboratory (GFDL) A1F1 climate scenario. We find that in situ and modeling methods are complementary, particularly for simulating historical and future recharge scenarios, and the in situ data are critical for accurately estimating recharge under current conditions. Observed (2011–2012) and future (2099–2100) recharge rates beneath the infiltration trench (1750–3710 mm yr−1) were an order of magnitude greater than beneath the irrigated lawn (130–730 mm yr−1). Beneath the infiltration trench, recharge rates ranged from 1390 to 5840 mm yr−1 and averaged 3410 mm yr−1 for El Niño years (1954–2012) and from 1540 to 3330 mm yr−1 and averaged 2430 mm yr−1 for La Niña years. We demonstrate a clear benefit for recharge and local groundwater resources using LID BMPs.