Noninvasive bone-conducted hearing aids require the consistent fitting of transducers for consistent hearing. In this paper, a method of estimating the contact force and amount of hair between a person’s skin and a bone-conducted sound transducer using electrical impedance is proposed. Experiments are conducted with a human surface model consisting of hair, skin, and bone. The estimator is implemented with a three-layered neural network. Ten measurements for 70 conditions are conducted by changing the contact force from 0 to 5 N and the amount of hair from 0 to 169.6 mm3. With the trained estimator, it is possible to estimate contact force and the intermediate material thickness with mean errors of 0.025 N and 0.424 mm3. This result supports the feasibility of the proposed method and contributes to the reproducible placement of the bone-conducted sound transducer.