The morphology of load-bearing tissues suggests that their internal void spaces fall into two different size distributions: macrovoids, the spaces between large morphological entities of several y in size, and microvoids, the spaces between structural elements. A mathematical model proposed by Lake and Armeniades gives an estimate of the fraction (F) of macrovoids in a given tissue on the basis of the rate of out-diffusion of ionic species from the tissue under controlled conditions. Knowledge of F would provide information on the state of aggregation of fibers or other structural elements in the tissue. This work seeks to determine whether the proposed model gives a true representation of the out-diffusion behavior in real systems with different pore size distributions. An inert porous material (Mi11ipore filter material type AAWP) was selected which contained two pore size populations. The properties of the material which correspond to the parameters of the model were observed by scanning electron microscopy. All parameters were quantified except D1, the tracer diffusivity in the microvoid region, and D2, the tracer diffusivity in the macrovoid region. D1 and D2 were varied to provide an exact fit of the model to the out-diffusion data. Variations of the parameters of the mathematical model show that the out-diffusion curve can be matched by only one unique value of F, regardless of how well the other parameters of the model are known. Out-diffusion experiments were carried out on tissue. The qualitative out-diffusion behavior of bovine aorta and pericardium suggests that the out-diffusion model may be used to obtain F with confidence. Variation in F and the sample half-thickness (M) for both tissues produces qualitative variation in the out-diffusion behavior which is predicted from the model.