We discuss a formalism for clinical proton beam dosimetry based on the use of ionization chamber absorbed dose-to-water calibration and beam quality correction factors. A quantity k(Q), the beam quality correction factor, is defined which corrects the absorbed dose-to-water calibration factor N-D,N-w, in a reference beam of quality Q(0) to that in a user's beam of quality Q(1). This study of proton beam quality correction factors used Co-60 (k(Q gamma)) and proton (k(Qp)) reference beams. The k(Q gamma) factors were measured using combined water calorimetry and ionometry for PTW and Capintec-Farmer-type ionization chambers, and were computed from standard dosimetry protocols. Agreement between measured and calculated k(Q gamma), values for both chambers was found within 1.2% in the plateau region for a monoenergetic 250-MeV beam and within 1.8% at the spread-out Bragg peak for a 155-MeV range-modulated beam. Comparison of absorbed doses to water determined in the range modulated 155-MeV beam was performed with the PTW chamber using three calibration methods: N-gas calibration (AAPM Report 16), N-D,N-w,N-gamma calibration in a Co-60 beam in conjunction with a k(Q gamma) factor, and N-D,N-w,N-p calibration in a proton beam in conjunction with a k(Qp), factor. Absorbed doses to water obtained with the three methods agreed within 2% when ionization chamber dosimetry data were analyzed using the proton W-value for air from the AAPM Report 16 and the ICRU 49 proton stopping powers. The use of the proton-calibrated reference ionization chamber, in conjunction with the beam quality correction factor k(Qp), significantly reduced the systematic uncertainty of the absorbed dose determination. (C) 1996 American Association of Physicists in Medicine.