We investigate the sources of variance in crop output and measure their relative importance in the context of weather index insurance for smallholder farmers. We use parcel-level panel data from South Asia and a multilevel modeling approach to isolate the different sources of variance. We then measure how large a role weather plays in explaining variance in yields. Using Bayesian methods, we draw the underlying distribution of the random error term responsible for weather uncertainty, which is highly skewed and non-normal. We find that variance in weather accounts for a small but important fraction of total variance in crop output. We also derive pricing and payout schedules for actuarially fair weather index insurance. Our results shed light on the low uptake rates of index insurance in South Asia and provide direction for designing index insurance with less basis risk for farmers.