SUMMARY
This review provides a survey of those factors that might influence genetic stability of genetically modified organisms and somatic hybrids in breeding programmes. In this respect several aspects may be distinguished: (i) host genomic factors that might influence genetic stability, (ii) events related to the introduction of new DNA into the genome and their effect on genetic stability, (iii) stability of gene expression of newly introduced DNA, and (iv) stability of the modified genome. In our view a gene is defined as being stable if it inherits according to Mendelian laws. Obviously, this can be valid only for nuclear genes. Non‐Mendelian inheritance may be caused by intrinsic genomic factors or be the result of skewed segregation during meiosis. Newly introduced DNA may be stably integrated into the genome, yet data on its site of integration is limited. The level of expression and, thus, the strength of the related trait, may vary. Variation in expression may depend on the construct, such as the promoter or additional sequences such as MAR elements or the coding sequence itself, the site of integration and the species used. Another, and undesired, phenomenon is the silencing of expression of introduced genes. The kinds of silencing described depend on the relative position in the genome of the genes involved, cis vs. trans and whether only one or all genes are silenced. Instability of expression generally becomes visible within a few generations, but once expression is stable it is supposed to remain so provided the environment does not change dramatically. Although the production of somatic hybrids seems to be a promising technique to obtain new genetic material and even though numerous hybrids have been made, only a few follow‐up studies have been published. Therefore the use of somatic hybrids in breeding programmes is limited.