Two male sterile lines (Ms-1, Ms-2) and ten inbred lines (K3, K4, K5, K8, K13, K15, K17, K20, K30, K33) were crossed according to Comstock and Robinson's (in: Gowen (ed.) Heterosis, 1952) North Carolina Design II to estimate heterosis, combining ability, heritability and the phenotypic correlation coefficient for 11 ornamental characteristics of Tagetes patula. Twenty cross combinations and 12 parents were evaluated for 11 ornamental characteristics, in a randomized complete block design with three replications, in trials conducted during 2011 and 2012 at Huazhong Agricultural University, Wuhan, China. The results showed that the hybrid combinations displayed higher ornamental values and showed obvious heterosis over the male-parent for most of the ornamental traits evaluated. The male sterile line Ms-1 scored highly for general combining ability (GCA) with respect to the development of F 1 offspring with a dwarf and compact plant habit, abundant double petals, and extended period between the opening and wilting of individual inflorescences. The male sterile line Ms-2 scored highly for GCA with respect to the development of F 1 offspring with a large plant form and large flower size, as well as abundant flower numbers. The specific combining ability (SCA) of cross Ms-2 9 K20 was found to be high, with progeny showing high positive heterosis over the male parent and superior mean values for all ornamental traits, thereby indicating that this cross has the potential to function as a superior cultivar within commercial breeding programmes. Both GCA effects and SCA effects were found to be important for the performance of individual hybrid combinations, and no clear relationship between GCA and SCA was found for these T. patula lines. The ratio of r 2 GCA =r 2 SCA for individual characteristics revealed that the relative importance of GCA and SCA to progeny quality varied for different traits. Estimates of heritability showed that many of the quantitative ornamental traits of T. patula were highly heritable, and the traits of diameter of center disk and number of ray florets should readily be improved by selection. By contrast, environmental factors influenced markedly on the traits of plant height, crown size, number of days between opening and wilting of inflorescences and the maximum number of flowers present each day. Phenotypic correlation coefficient analysis showed that the various vegetative growth characteristics were closely related to one another and, more interestingly, significant phenotypic correlations were found between the growth of the vegetative and reproductive structures of T. patula.