A field experiment conducted in the field of Crops Science Dept., Coll. of Agric. Univer. of Baghdad, during 2001-2002 and 2002-2003, seasons. The objectives was to study the path analysis in bread wheat traits. The experiments were carried out, using split plot design with ten varieties (Tamoz 2, Tamoz 3, Abu-Ghraib 3, IPA 95, IPA 99, Sale, Al-Rashed, Al-Iraq, Fateh and rabea) as main plots and four seeding rates (80, 120, 240 and 320 kg / ha) as sub plots with three replications. Path analysis revealed that the harvest index for all seeding rates, biological yield of higher seeding rates and number of spikes / m2 in low seeding rates, obtained highest direct effect to grain yield.The highest direct and indirect effect to grain yield revealed from harvest index, spike length and number of spikelets / spikes.It was cocluded that harvest index could be used as selection index for the improvement of grain yield in bread wheat.