A controllable trunk shaker was modeled as part of the design process for construction of a shaker that would not cause visible or nonvisible bark damage. Two different vibrational models were employed. A single degree-of-freedom model was used to optimize the size and shape of the shaker eccentric masses. A multiple degree-of-freedom model was developed to predict shaker behavior when shaking trees of various trunk diameters. Field tests conducted with the constructed shaker indicated qualitative agreement with the model was good. Quantitative agreement was also satisfactory, with peak-to-peak amplitudes within 30% of actual values.