The verification of the correctness of planned and executed treatments is imperative for safety in radiotherapy. The purpose of the present work is to describe and evaluate the quality assurance (QA) procedures for patient dosimetry implemented at the boron neutron capture therapy (BNCT) facility at Studsvik, Sweden. The dosimetric complexity of the mixed neutron-photon field during BNCT suggests a careful verification of routine procedures, specifically the treatment planning calculations. In the present study, two methods for QA of patient dosimetry are presented. The first is executed prior to radiotherapy and involves an independent check of the planned absorbed dose to be delivered to a point in the patient for each treatment field. The second QA procedure involves in vivo dosimetry measurements using post-treatment activation analysis. Absorbed dose conversion factors taking the difference in material composition and geometry of the patient and the PMMA phantom used for reference dosimetry were determined using the Monte Carlo method. The agreement of the QA procedure prior to radiotherapy reveals an acceptably small deviation for 60 treatment fields of +/-4.2% (1 SD), while the in vivo dosimetry method presented may benefit from improvements, as the deviations observed were quite substantial (+/- 12%, 1 SD), and were unlikely to be due to actual errors in the clinical dosimetry