A field trial was conducted in analluvial soil of Bangladesh to evaluate the effect of boron on the number of grains/spike, grain yield and nitrogen content of wheat sown in different dates. The experiment comprised two B treatments (O and 3 kg B/ha), three wheat varieties (Sonalika, Kanchan and Aghrani) and four dates of sowing with 2-weeks intervals beginning from the November 15 of 1992. The experiment was laid out in split-split-design with sowing date as main-plot, wheat variety as sub-plot and boron treatment as sub-sub-plot. The results showed that added B had positive effect on the number of grains/spike and grain yield. Both grains/spike and grain yield were decreased as the date of sowing was delayed particularly after December 1, over the varieties. Among the varieties, Kanchan gave comparatively better yield. Nutrient analysis showed that N and B contents in grain were increased after B treatment to soil indicating that B probably helped protein synthesis. Like grain yield, grain N level had declined with the delaying of sowing date. Such adverse effect of late sowing was thought to be related with elevated temperature prevailing at flowering stage that induced slower rate of B mobility in plant, poor number of grains/spike and lower grain yield.