Two inherent difficulties have been responsible for the relatively neglected state of cytoplasmic inheritance in experimental genetics. These are the apparent scarcity of cytoplasmic differences and the difficulty of proving the cytoplasmic origin of differences in phenotype. Both difficulties have been overcome by utilizing the natural advantages of micro-organisms and more especially those of the heterokaryotic fungi. Both natural and artificial heterokaryons can readily be made in the laboratory with many of the filamentous ascomycetous fungi, and we have used three such fungi, namely Penicillium cyclopium , Aspergillus glaucus and A. nidulans . In the heterokaryotic state an almost indefinite co-existence in the same cytoplasm of nuclei from different individuals can be achieved. This property, along with the ease with which the nuclei can be re-extracted from the heterokaryon by asexual spores, allows an unambiguous classification of the differences between two individuals into those of nuclear and cytoplasmic origin. The details of this method of separation, which is described by Jinks (1954, 1956), are summarized in table 1. Where both nuclear (other than the introduced marker) and cytoplasmic differences are responsible for differences between two homokaryons the situation can be more complex than that given in table 1. No general account of the consequences of asexual and sexual extraction of two such components from a heterokaryon is therefore possible; each case may show its own peculiarities and complications depending on the nature and extent of any nuclear-cytoplasmic interaction in the heterokaryotic state. However, as we shall see later, such examples have been successfully dealt with by this approach, which thus covers all the theoretically possible differences between two phenotypes.