Heterosis, or hybrid vigour, describes the superior performance of heterozygous hybrid individuals compared with their homozygous parental inbred lines. Although heterosis has been intensively used in plant breeding, the molecular and genetic mechanism underlying the phenomenon remains largely unknown. During the last years numerous laboratories initiated genomic approaches with new, often genome wide tools toward the elucidation of the molecular basis of heterosis. Various studies described differences in genome organization and gene expression of hybrids and their parental inbred lines. In maize, a considerable loss of co-linearity at many loci between different inbred lines was observed. Expression profile comparisons between inbred lines and hybrids revealed complex transcriptional networks specific for different developmental stages and tissues mainly in maize (Zea mays), rice (Oryza sativa) and Arabidopsis (Arabidopsis thaliana). Allele-specific expression data resolved the relative parental contributions and allowed figuring out the regulatory basis for expression variation in hybrids. Integrating all these complex expression data might help to get an idea about the molecular basis of heterosis. Thereby molecular processes during early seed development shortly after fertilization might be of particular importance, because the allelic interplay has to be coordinated after the unification of two diverse genomes and these processes might contribute to establish the basis for future performance of the sporophyte. Besides these fundamental interests in the molecular basis and manifestation of hybrid vigour applied aspects of the phenomenon are of high importance to support plant breeding and agriculture. Prediction methods are of special interest to identify the most promising parental lines of hybrid varieties, greatly reducing the financial effort and increasing the efficiency to develop new hybrid cultivars. Until now, most prediction approaches were based on genomic markers. The available heterosis associated expression data provide evidence to support the idea that gene regulatory networks at the level of transcription are involved in the control of hybrid vigour. If so, the transcriptome characteristics of inbred lines resulting from their individual genomic constitution should be useful as quantitative markers to predict the performance of hybrids generated by crosses of these inbred lines. Additionally, expression based prediction models promise to contribute substantially to the understanding of the genetic causes of heterosis by establishing a direct link between the transcriptome and hybrid performance. Heterosis, or hybrid vigour, describes the superior performance of heterozygous parental inbred lines. Although heterosis has been intensively used in plant breeding, the molecular and genetic mechanism underlying the phenomenon remains largely unknown. During the last years numerous laboratories initiated genomic approaches with new, often genome wide tools toward the elucidation of the molecular basis of heterosis. Various studies said differences in genome organization and gene expression of hybrids and their parental inbred lines. In maize, a considerable loss of co-linearity at many loci between different inbred lines was observed. Expression profile comparisons between inbred lines and hybrids revealed complex transcriptional networks specific for different developmental stages and tissues mainly in maize (Zea mays), rice (Oryza sativa) and Arabidopsis (Arabidopsis thaliana). Allele-specific expression data resolved the relat ive parental contributions and allowed figuring out the regulatory basis for expression variation in hybrids. Yet molecular processes during early seed development shortly after fertilization might be of particular importance Because the allelic interplay has to be coordinated after the unification of two diverse genomes and these processes might contribute to establish the basis for future performance of the sporophyte. are of high importance to support plant breeding and agriculture. Prediction methods are of special interest to identify the most promising parental lines of hybrid varieties, greatly reducing the financial effort and increasing the efficiency to develop new hybrid cultivars. Until now, most prediction approaches were based on gThe available heterosis associated expression data data provide evidence to support the idea that gene regulatory networks at the level of transcription are involved in the control of hybrid vigour. If so, the transcriptome characteristics of inbred lines resulting from their individual genomic constitution should be useful as quantitative markers to predict the performance of hybrids generated by crosses of these inbred lines. Additionally, expression based prediction models promise to contribute substantially to the understanding of the genetic causes of heterosis by establishing a direct link between the transcriptome and hybrid performance.