In the alkane yeast Saccharomycopsis lipolytica (formerly: Candida lipolytica) the variability in the ascospore number is caused by the absence of a correlation between the meiotic divisions and spore wall formation. In four spored yeasts, after meiosis II, a spore wall is formed around each of the four nuclei produced by meiosis II. However, in the most frequently occurring two spored asci of S. lipolytica, the two nuclei are already enveloped by the spore wall after meiosis I due to a delay of meiosis II. This division takes place within the spore during the maturation of the ascus. In this case germination of the binucleate ascospore is not preceded by a mitosis. It follows that the cells of the new haploid clones are mononucleate. In the three spored asci, which occur rarely, only one nucleus is surrounded by a spore wall after meiosis I; the other nucleus undergoes meosis II before the onset of spore wall formation. The result is one bincleate and two mononucleate spores. In the one spored asci the two meiotic divisions occur within the young ascosphore, i.e. spore wall formation starts immediately after development of the ascus. These cytological observations were substantiated by genetic data, which in addition confirmed the prediction that binucleate spores may be heterokaryotic. This occurs when there is a postreduction of at least one of the genes by which the parents of the cross differ. This also explains the high frequency of prototrophs in the progeny of non-allelic auxotrophs since random spore isolates are made without distinguishing between mono- and binucleate spores. The possibility of analysing offspring of binucleate spores by tetrad analysis is discussed. These findings enable us to understand the life cycle of S. lipolytica in detail and we are now in a position to start concerted breeding for strain improvement especially with respect to single cell protein production.