Overview and scope of the presentation: Efficient exploitation of heterosis is central for successful hybrid breeding in sunflower. Here we discussed important genetic hypotheses of the phenomenon of heterosis including dominance overdominance and epistasis. Moreover we presented different approaches suggested for prediction of hybrid performance. Outline of recent major discoveries: Several experimental designs have been proposed to study the causes of heterosis. Among them QTL analyses based on triple test cross in combination with the immortalized F2 design or the use of triple test cross designs based on near-isogenic lines are of particular interest. Several approaches have been used to predict hybrid performance using phenotypic genomic transcriptomic and metabolomic data. Prediction accuracies have been studied using experimental data of maize which clearly points towards the large potential for knowledge-based hybrid breeding. Perceived significant gaps in research current debates and perspectives for future directions of research: Experimental design and quantitative genetic theory is ready to be applied in sunflower to study the genetic basis of heterosis. Furthermore we speculate that genomic selection possesses great potential to revolutionize hybrid prediction in sunflower.