The purposes of this study were to investigate the influence of the consequences of the lack of primary bioassay information and to elaborate approaches which could improve the reliability of dose assessments. The aggregated time-dependent functions 'dose per unit organ (excretion) content' z(t) have been proposed in this study as a convenient and reliable tool for bioassay. The analysis of the variation of z with changes of AMAD has demonstrated the existence of areas of the relative invariance of z, which permits the selection of one (reference) function z for the whole area of stability. Within the framework of such an approach an arbitrary set of bioassay data can be approximated by the linear combination F(t) = sum(i) E(i)/z(t - tau(i)), where F(t) = function of time t, which approximates the observed bioassay time trend; tau(i) = time shift of the acute intake i; E(i) = effective dose, associated with the acute intake i (the two last parameters are results of the approximation procedure).