To evaluate the daily setup accuracy and the reduction of respiratory tumor movement using a body frame in conformal therapy for solitary lung tumor. Eighteen patients with a solitary lung tumor underwent conformal therapy using a body frame. The body shell of the frame was shaped to the patient's body contour. The respiratory tumor movement was estimated using fluoroscopy, and if it was greater than 5 mm, pressure was applied to the patient's abdomen with the goal of minimizing tumor movement. CT images were then obtained, and a treatment planning was made. A total dose of 40 or 48 Gy was delivered in 4 fractions. Portal films were obtained at each treatment, and the field displacements between them and the simulation films were measured for daily setup errors. The patients were repositioned if the setup error was greater than 3 mm. Correlations were analyzed between patient characteristics and the tumor movement, or the tumor movement reduction and the daily setup errors. Respiratory tumor movement ranged from 0 to 20 mm (mean 7.7 mm). The abdominal press reduced the tumor movement significantly from a range of 8 to 20 mm to a range of 2 to 11 mm (p = 0.0002). Daily setup errors were within 5 mm in 90%, 100%, and 93% of all verifications in left-right, anterior-posterior, and cranio-caudal directions, respectively. Patient repositioning was performed in 25% of all treatments. No significant correlation was detected between patient characteristics and tumor movement, tumor movement reduction, and the daily setup errors. The abdominal press was successful in reducing the respiratory tumor movement. Daily setup accuracy using the body frame was acceptable. Verification should be performed at each treatment in hypofractionated conformal therapy.