Purpose:To evaluate a patient‐specific QA program and system for constancy checking of a scanning delivery system developed at the National Institute of Radiological Sciences.Methods:For the patient‐specific QA, all the planned beams are recalculated on a water phantom with treatment planning software (TPS). The recalculated dose distributions are compared with the measured distributions using a 2D ionization chamber array at several depths, and evaluated using gamma index analysis with criteria of 3% and 3 mm and a pass rate of 90%. For the constancy check, the authors developed the multiwire proportional chamber (MWPC), which can record the delivered 2D fluence images in a slice‐by‐slice manner. During irradiation for dosimetric QA with the 2D ionization chamber array and an accordion‐type water phantom, the 2D fluence images are recorded using the MWPC in the delivery system. These recorded images are then compared to those taken in the treatment session to check the constancy check. This analysis also employs gamma index analysis using the same criteria as in the patient‐specific QA. These patient‐specific QA and constancy check evaluations were performed using the data of 122 patients.Results:In the patient‐specific QA, the measured dose distributions agreed well with those calculated by the TPS, and the QA criteria were satisfied in all measurements. The additional check of the fluence comparison ensured the constancy of the delivered field during each treatment irradiation.Conclusions:The authors established a patient‐specific QA program and additional check of delivery constancy in every treatment session. Fluence comparison is a strong tool for constancy checking of the delivery system.