Grain yield in bread wheat (Triticum aestivum L.) is influenced by root and shoot characteristics and phenological periods. Our objective was to determine optimum root and shoot biomass to maximize grain production. Genotypes included 12 recombinant inbred lines (RILs) derived from a cross Iran #49 × Yecora Rojo (YR) plus two parents, and two California wheats, ‘Blanca Fuerte’ (BF) and ‘Joaquin’. Genotype #49 is a tall late landrace with large root system, YR is a CIMMYT-derived two-gene dwarf with small root system grown locally. Seven days-old seedlings with similar growth were transplanted in polyethylene tubing bags sleeved into polyvinyl chloride (PVC) tubes, 150 cm long and 10 cm diameter containing 16 kg dry silica sand with 24% field capacity in a glasshouse at University of California, Riverside, on January 20, 2012. A randomized complete block design with four replicates was used. Plants were grown under well-watered, fertilized conditions. Significant differences were found among the genotypes for all the traits measured including root biomass (RB), shoot biomass excluding grains (SBEG), grain yield (GY), and harvest index (HI). Grain yield was not correlated with days to maturity or plant height. Two hyperbola curves determined the relationship between GY with RB (R2 = 0.93) and with SBEG (R2 = 0.88). The curves indicated maximum grain yield (16.4 g plant-1) was produced when RB and SBEG reached 9 and 22 g plant-1, respectively. Negative correlation existed between RB and HI. Root biomass in YR, Joaquin, and BF was 1.92, 1.93, and 3.06 g plant-1, #49 was 9.48 g plant-1. Root biomass among the RILs varied from 0.92 to 11.76 g plant-1. One RIL had GY and HI similar to BF, the best check, but with greater RB. Plants with larger root system might also be more adapted to rain fed conditions.