A model that predicts the osmotic fragility curve of a red cell population is developed by relating the critical osmotic pressure to the size distribution of the cells, determined by resistive pulse spectroscopy. Two of the parameters involved, namely the normalized osmotic volume correction, B, and the swelling index, k, are previously determined from the experimental average properties of the population. From these values the critical volume of the cell is obtained, and is shown to be 6-12% larger than the first spherical volume, obtained from an independent experiment. A new parameter, n, a measure of the surface area distribution of the cells, is incorporated through a simple function that relates the critical volume to the size of the cells, and is theoretically shown to be linked to parameters k and B. The model is used to fit and interpret fragility data obtained in this laboratory for normal and sickle cell samples. From the values of n obtained for normal samples, the model predicts an essentially constant surface-to-volume ratio within an individual's cell population. For sickle cell samples, instead, the value of index n is negative, thereby supporting an increase in excess surface area as cell size decreases. Both findings are in agreement with direct observations reported in the literature. It is concluded that this set of parameters may be used to develop an index classification of blood disorders.