Hourly end-use load information is extremely useful to utilities for purposes of system planning and forecasting, demand-side management, and peak-load planting. Load data transfer - borrowing data from other service territories and/or time periods – is less expensive than direct metering. However, not much is known about the imprecision and statistical bias introduced by methods of load data transfer. We evaluate the accuracy of 11 low-cost load data transfer methods for residential air conditioner use. We use each method to predict load shapes which are then compared to a shared set of actual end-use measured loads. We conclude that the degree of imprecision and bias introduced by each method can be quantified, at least in a preliminary way, and that low-cost methods like the ones we evaluated may be cost-effective for many purposes for which utilities use end-use load data