In India, herbicide use has almost tripled in the decade from 2005/2006 to 2015/2016, while cultivated land did not change over that period. This is a striking case of input intensification, driven by rapidly rising rural wages, combined with soaring domestic production of herbicides, and modestly declining herbicide prices. Rural wages are rising, in part due to the rapid growth of rural nonfarm employment, and the tightening opportunity costs of time for rural Indians to manually weed their farms. To explore the household-level determinants and effects of herbicide adoption, we used a unique primary survey dataset of 2418 farm households collected by the authors in 2009 in the more commercialized regions (mainly growing cereals and pulses) over the east, center, and west of each of Uttar Pradesh (UP), Madhya Pradesh (MP), and Andhra Pradesh (AP). These three states range from north to south, and include 330 million people. The survey’s descriptive analysis showed that herbicide diffusion is well advanced, covering one third of farmers in UP and AP, and two-thirds in MP. The share had jumped 1.5, 1.2, and 2.7 times over just five years in the three states (UP, MP, and AP, respectively). The diffusion was broader and faster in the more dynamic regions of the states – in cereals and horticulture compared with legumes – and in larger farms in UP and MP (but in smaller farms in the irrigated rice area of AP). The regression analysis of adoption showed that as the rural nonfarm employment increases, the probability of adoption of herbicides increases very significantly. Herbicide use boosts land and labor yields by about 10%, and reduces labor use by approximately 5–10%. Agricultural research should be oriented toward herbicide formulations, which makes them adaptable to evolution of weed and agroclimatic conditions, and as affordable and as safe as possible for workers and the land.