Abstract A statistical examination of isotopic distributions for MORB from various ocean ridges leads to the “blob cluster model”, in which the oceanic crust accreting at ridges results from the mixing of two components within the ascending mantle. These are (1) upper mantle material and (2) discrete rising blobs of more radiogenic material. The blobs are fractionated to a variable degree and are distributed in the upper mantle circulation in a manner that is related to the spreading rate. (1) The mean values of the isotopic distributions allow us to calculate the probabilities of the two types of material within the mantle. The results show that the proportion of asthenospheric material in the mixture increases with the spreading rate, in agreement with the hypothesis of blob dilution within the upper mantle convection. Mass fluxes can be estimated for the rising blobs from these probabilities, which depend on the respective concentrations in the sources of the two types of material. If the blobs originate in the lower mantle, this flux estimation would suggest that a significant part of the lower mantle has been injected into the upper mantle during earth history. (2) The standard deviations of the distributions depend on the “efficiency” of the mixing process: the more imbricated are the asthenospheric and blob materials in the mixture, the smaller is the isotopic spread. This efficiency parameter is shown to increase with the spreading rate, as already suggested by previous comparisons between the East Pacific Rise and the Mid-Atlantic Ridge. Moreover, this feature may also be correlated with other data such as ridge bathymetric variations.