At the present time, there is no easily performed method of measuring foot shape that can quantitatively differentiate types of feet. We studied 44 feet of individuals with normal appearing asymptomatic feet and flat symptomatic feet to find criteria for measurement of foot shape. We found that measurements of footprints are unreliable. Radiographic methods must include measurements in different positions, or at least specify the position of subtalar rotation to be meaningful, since the height of the arch can change just by rotating the tibia. Traditional measurements of talocalcaneal angles have no precision and are unreproducible.