The dominance of biofuel combustion emissions in the Indian region, and the inherently large uncertainty in biofuel use estimates based on cooking energy surveys, prompted the current work, which develops a new methodology for estimating biofuel consumption for cooking. This is based on food consumption statistics, and the specific energy for food cooking. Estimated biofuel consumption in India was 379 (247--584) Tg yr-1. New information on the user population of different biofuels was compiled at a state level, to derive the biofuel mix, which varied regionally and was 74:16:10%, respectively, of fuelwood, dung cake and crop waste, at a national level. Importantly, the uncertainty in biofuel use from quantitative error assessment using the new methodology is around 50%, giving a narrower bound than in previous works. From this new activity data and currently used black carbon emission factors, the black carbon (BC) emissions from biofuel combustion were estimated as 220 (65--760) Gg yr-1. The largest BC emissions were from fuelwood (75%), with lower contributions from dung cake (16%) and crop waste (9%). The uncertainty of 245% in the BC emissions estimate is now governed by the large spread in BC emission factors from biofuel combustion (122%), implying the need for reducing this uncertainty through measurements. Emission factors of SO2 from combustion of biofuels widely used in India were measured, and ranged 0.03--0.08 g kg-1 from combustion of two wood species, 0.05--0.20 g kg-1 from 10 crop waste types, and 0.88 g kg-1 from dung cake, significantly lower than currently used emission factors for wood and crop waste. Estimated SO2 emissions from biofuels of 75 (36--160) Gg yr-1 were about a factor of 3 lower than that in recent studies, with a large contribution from dung cake (73%), followed by fuelwood (21%) and crop waste (6%).