. 65, 1079 (1987). The efficiency of an HGA-400 for production and containment of atomic vapour was determined for several elements using relatively high heating rates. Two different atomization efficiencies P I and P2 were experimentally determined by two variations k l k, of a method. Assuming that atomization in the HGA-400 can be described by consecutive, first-order reactions D -B t C, the method consisted of making the absorbance signal profile (at least the decay part of it) appear at constant tube wall (at the centre) temperature and determining first-order rate constants for atom formation and atom loss, k l and k 2 , respectively.Two atomization efficiencies were calculated: P I and p 2 , which gave the fraction of analyte which was present as atomic vapour at the peak of the absorbance signal profile. Experimental atomization efficiencies reported for Al, Cu, Co, Fe, Ag, and Ga average: p 0.42 and p2 0.30, with a range from 0.30 to 0.50 for P I and from 0.18 to 0.43 for p 2 . Significance of the results and limitations of the two variations of the method in terms of validity of their underlying assumptions have been brought out in a critical assessment of the atomization efficiencies measured by the two variations of the method. In both the variations of the method, the heating rates used were much higher than those used by some other authors, with the result that in agreement with the prediction of the theory, the atomization efficiencies measured in this work were considerably higher than those reported by those authors. [Traduit par la revue]