Transfer factor (TL) varies with alveolar volume (VA), but not in the manner implied by the carbon monoxide transfer coefficient (KCO (TL/VA)). This paper considers two other simple models (one linear and one exponential) which might standardize TL for VA, and asks the questions: 1) Is either model valid? 2) What are appropriate reference values? and 3) Will the model be useful? The relationship of TL to VA within subjects at different depths of inspiration, and between subjects having lungs of different sizes, were measured and compared. The subjects were asymptomatic, nonsmoking, Caucasian adults, including 31 males assessed in the laboratory and 503 male and female participants in population studies. The linear partial regression coefficients of TL on VA (L corrected for body temperature, atmospheric pressure and water saturation (BTPS)) standardized for height (H) in metres, were similar within- and between-subjects; the coefficients applied over a wide range of values for VA. This was not the case for the exponential model. The resulting reference equations in SI units for males and females were: TL = 11.52 H + 2.72 VA.H-2 - 0.051 Age -12.35. RSD 1.17; and TL = 4.87 H + 2.29 VA.H-2 - 0.019 Age -3.03. RSD 0.92, respectively. The residual standard deviations (RSD) about the new relationships were less than in other series. The new linear model could account for much of the variation between different published reference values for TL; it could be useful clinically, in circumstances when VA deviates from the norm. The model does not explain differences in TL associated with gender. Inclusion of VA.H-2 as a covariate in the reference equation for transfer factor, in addition to age and height, improves the accuracy of prediction of normal transfer factor compared with current reference values; its use suggests that some of the differences between published values is due to the volume term. The equations can be used clinically, and eliminate the need for carbon monoxide transfer coefficient.