The investigation was performed using six parental genotypes and resultant 15 half-diallel crosses in sorghum. The analysis of variance for combining-ability (CA) studies revealed that mean sum of squares due to general-combining-ability (GCA) were extremely important for most of the characters except the leaves number plant^-1. The unit slope of the regression line and nonsignificant t² value asserted the validity of the model of additive-dominance for flowering-maturity days, plant height, leaves number plant^-1, stem girth, leaf width, green fodder yield plant^-1, dry fodder yield plant^-1 and brix value. The value of H₁ was superior to H₂ for most of the characters representing that gene sharing frequency in the parent genotypes was not equal and that was also supported by the ratio of H₂/4H₁ (i.e., less than 0.25). The estimations of F element was positive and nonsignificant, while of KD/KR components value was greater than one among most of the traits except for plant height and leaf width showed that parents had more dominant genes compared to recessive genes frequency. In Hayman graphical investigation, the regression line interrupted Wr axis underneath the origin demonstrating over-dominance for flowering days, height of plant, leaves number plant^-1, girth of stem, leaf width, green fodder yield plant^-1, dry fodder yield plant-1 and brix value.