Purpose: This study aims to devise and identify the errors of the beam width measurements in a wide beam CT system using three tungsten rings (TR) in comparison with a flat-panel detector (FPD), and develop a new method to correct the errors. Methods: A pencil-type ionization chamber was placed at the isocenter. The Kerma-length product ( KLP) was measured at 80 and 120 kV, 400 mA, and a rotation time of 0.5s with nominal beam widths ranging from 20 to 160 mm in 20 mm increments. Subsequently, each TR was attached to the chamber to measure the KLP_mask at the isocenter, and the beam width was calculated as KLP×ring length/( KLP−KLP_mask). To compare the measurement accuracy, the beam widths were measured using the FPD with a double-exposure technique. The X-ray exposures were performed at 80 kV, rotation time of 0.5 s, and 10 and 20 mA were used for the measurements. Finally, the heel effect correction, replacing the KLP at the anode side, was also compared. Results: The measured beam widths using 5-, 10-, and 15-mm TRs at 80/120 kV, and the FPD were 182.5 /182.1, 167.5/165.7, 168.2/163.0, and 172.9 mm in the nominal beam width of 160 mm, respectively. The heel effect correction with 10- and 15-mm TRs at 80 kV improved the measurement accuracy, and the corrected beam widths were 172.4 and 173.2 mm, respectively. Conclusion: In conclusion, 10- and 15-mm TRs in conjunction with the heel effect correction are appropriate for the beam width measurements in a wide beam CT system.