We understand little about the energetic costs of flight in free‐ranging birds, in part because current techniques for estimating flight energetics in the wild are limited. Accelerometry is known to estimate energy expenditure through body movement in terrestrial animals, once calibrated using a treadmill with chamber respirometry. The flight equivalent, a wind tunnel with mask respirometry, is particularly difficult to instigate, and has not been applied to calibrate accelerometry. We take the first steps in exploring a novel method for calibrating accelerometers with flight energy expenditure. We collected accelerometry data for Harris's Hawks Parabuteo unicinctus flying to varying heights up to 4.1 m over a small horizontal distance; the mechanical energy expended to gain height can be estimated from physical first principles. The relationship between accelerometry and mechanical energy expenditure was strong, and while a simple wing flapping model confirmed that accelerometry is sensitive to both changes in wing beat amplitude and frequency, the relationship was explained predominately by changes in wing beat frequency, and less so by changes in amplitude. Our study provides initial, positive evidence that accelerometry can be calibrated with body power using climbing flights, potentially providing a basis for estimating flapping flight metabolic rate at least in situations of altitude gain.