An experiment was conducted at Research Station, State Board for Seeds Testing and Certification, Baghdad, Iraq during 2015-2016 and 2016-2017 seasons to determine the effect of  four sowing dates (20^th October, 10^th November, 30^th November and 20^th December) on grain yield and it's components of three wheat cultivars (Bohooth 22, Bohooth 158 and Rasheed). The experiment was laid out in Randomized Complete Blocks Design (RCBD) within split plot arrangement. Sowing dates distributed in main plots and cultivars in sub plots with three replications. Sowing date 10^th November was superior in grain yield (4.96 and 5.21 t ha^-1) and most of its components (number of spikes 403.3 and 408.6 spike m^-2 and number of spikelets per spike 22.08 and 22.26 spikelet spike^-1 and number of grains per spike 56.37 and 58.63 grain spike^-1), and some other traits (number of tillers 448.6 and 442.0 tiller m^-2 and spike length 15.84 and 14.39 cm) for both seasons, respectively, whereas sowing in 20^th October was superior in 1000 grain weight (42.94 and 43.72 g) for both seasons, respectively. Cultivar Rasheed was superior in grain yield (4.50 and 4.71 t ha^-1) and most of yield components (number of spikes 389.5 and 397.3 spike m^-2 and number of spikelets per spike 22.52 and 22.59 spikelet spike^-1 and 1000 grain weight 41.61 and 43.02 g) and in some other traits (number of tillers 428.7 and 432.4 tiller m^-2 and spike length 17.27 and 16.69 cm) for both seasons, respectively. It's clear from the results that 10^th November is the best sowing date at central Iraq to get highest grain yield, and the early or late sowing could be decrease the grain yield, although the effect of early sowing is greater than the effect of late sowing on this trait in wheat.