The purpose of this study was to suggest tolerance levels for IMRT DQA measurements using confidence limits determined by a multi-institutional study in Korea. Ten institutions were grouped into LINAC (seven linear accelerators) and TOMO (three tomotherapy machines). The DQA processes consisted of point (high- and low-dose regions) and planar (per-field and composite-field) dose measurements using an ion chamber and films (or 2D detector array) inserted into a custom-made acryl phantom (LINAC) or a cheese phantom (TOMO). The five mock structures developed by AAPM TG-119 were employed, but the prostate as well as the H&N structures were modified according to Korean patients' anatomy. The point measurements were evaluated in a ratio of measured and planned doses, while the planar dose distributions were assessed using two gamma criteria of 2 mm/2% and 3 mm/3%. The confidence limit (|mean + 1.96 σ|) for point measurements was determined to be 3.0% in high-dose regions and 5.0% in low-dose regions. The average percentage of points passing the gamma criteria of 2 mm/2% and 3mm/3% for per-field measurements was 92.7 ± 6.5% and 98.2 ± 2.8%, respectively. Thus, the corresponding confidence limit was 79.1% and 92.7%, respectively. The gamma passing rate averaged over all mock tests and institutions for composite-field measurements was 86.1 ± 6.5% at 2 mm/2% and 95.3 ± 3.8% at 3 mm/3%, leading to the confidence limit of 73.3% and 87.9%, respectively. There was no significant difference in the tolerance levels of point dose measurements between LINAC and TOMO groups. In spite of the differences in mock structures and dosimetry tools, our tolerance levels were comparable to those of AAPM and ESTRO guidelines.