In this study, transplanting cotton seedlings grown in artificial substrate is considered due to recent increased interest in cotton planting labor saving approaches. The nursery methods used for growing cotton seedlings affect root growth. However, the underlying functional responses of root growth to variations in cotton seedling transplanting methods are poorly understood. We assessed the responses of cotton (Gossypium hirsutum L.) roots to different planting methods by conducting cotton field experiments in 2012 and 2013. A one-factor random block design was used with three replications and three different cotton planting patterns (substrate seedling transplanted cotton (SSTC), soil-cube seedling transplanted cotton (ScSTC) and directly sown cotton (DSC). The distributions and variances of the root area density (RAD) and root length density (RLD) at different cotton growing stages and several yield components were determined. Overall, the following results were observed: 1) The RAD and RLD were greatest near the plants (a horizontal distance of 0 cm) but were lower at W20 and W40 cm in the absence of film mulching than at E20 and E40 cm with film mulching. 2) The roots were confined to shallow depths (20-40 cm), and the root depths of SSTC and DSC were greater than the root depths of ScSTC. 3) Strong root growth was observed in the SSTC at the cotton flowering and boll setting stages. In addition, early onset root growth occurred in the ScSTC, and vigorous root growth occurred throughout all cotton growth stages in DSC. 4) The SSTC plants had more lateral roots with higher root biomass (RB) than the ScSTC, which resulted in higher cotton yields. However, the early onset root growth in the ScSTC resulted in greater pre-frost seed cotton (PFSC) yields. These results can be used to infer how cotton roots are distributed in soils and capture nutrients.