An experiment was carried out to study the performance of twenty seven onion accessions under different planting date during the rabi season of 2012-2013. The research work was carried out at the USDA Allium field laboratory, Horticulture Farm, Bangladesh Agricultural University, Mymensingh. The experimental area belongs to the Old Brahmaputra Flood Plain of AEZ 9 having non-calcareous Dark Grey Flood Plain soil (FAO, 2012). The experimental field was medium high land. It was fertile and well drained and slightly acidic with the pH varying from 5.5 to 6.8 (BARC, 2005). The experiment consisted of two factors, planting date viz. 20 December, 20 January and accessions viz. Onion-1, Onion-2, Onion-3, Onion-4, Onion-5, Onion-6, Onion-7, Onion-8, Onion-9, Onion-10, Onion-11, BAU AC-1, BAU AC-2, BAU AC-4, BAU AC-6, BAU AC-7, BAU AC-13, BAU AC-15, BAU AC-19, BAU AC-21, BAU AC-23, BAU AC-24, BAU AC-25, BAU AC-26, BAU AC-31, BAU AC-32, BAU AC-33. The experiment was laid out in a Randomized Complete Block Design (RCBD) with three replications. The area of unit plot was 1m x 1m. The spaces between blocks and between plots were 1.0 m and 50 cm, respectively. The experimental plot was opened in the month of October 2012 with the help of tractor. Thereafter, the land was prepared by several ploughings and cross ploughings with a power tiller followed by laddering and was uniformly fertilized with cow dung, urea, triple super phosphate, muriate of potash, gypsum and zinc oxide @ 10 t/ha and 200, 275, 150, 110 and 3 kg ha-1 respectively. The whole amount of well decomposed cow dung TSP, gypsum and zinc oxide was applied during the final land preparation. The total quantity of Urea and MP was applied as top dressing in two equal installments, first after 30 days and the second after 50 days of transplanting of seedlings (Rashid, 1983). On the other hand, the area for seedbed was spaded and left opens for few days. After removal of trashes and weeds, the soil was mixed with well rotten cattle manure. Finally, the land was prepared by mixing with recommended doses of manures and fertilizers. Irrigation and drainage channels were made around the plots. Seeds were soaked overnight in water as suggested by Lipe and Skinner (1979) and allowed to sprout in a piece of cloth keeping in the sunshade for two days. The first date of seed sowing was 28 October, 2012. The sprouted seeds (3-4 in number) were sown directly in the raised seedbed for raising seedlings to be transplanted later on as per treatments. Following the same procedure, the seeds were sown on the next date 28 November, 2012. Shades were given over the seedbeds to retain soil moisture and to save the seedlings from direct sun and rain. When the seedlings of the seedbeds attained a height of about 10 cm, thinning operation was done keeping only the healthy seedling in right place. Transplanting of seedlings was done on 20 December, 2012 and 20 January, 2013. At the age of 50-55 days, the seedlings were transplanted in the main field with the same spacing (15 cm x 10 cm). At the time of lifting the seedlings, care was taken to avoid any mechanical injury to the seedlings. Gap filling was done properly as early as possible in order to get uniform plants per plot. The intercultural operations were done whenever it necessary. The crops were harvested according to their attainment of maturity. Before 7 days of harvest, when the plants attained maturity by showing drying up of leaves and weakening of necks, the crop was bended at the soil level by hands and kept as such up to harvest to fasten maturity. The entire transplanting crops were harvested on 28 March to 1 April, 2013 for convenience in data recording. The harvested crop of each plot was separately bundled, properly tagged and then brought to the threshing floor. The bulbs were then cleaned properly. Finally the bulb yield plot-1 were recorded and converted to t ha-1. Data recorded on different parameters were compiled and tabulated in proper form. Analysis of variance was done following Randomized Complete Block Design with the help of computer package program MSTAT. The means for all the treatments were calculated and analyses of variances for all the characters were performed by F-variance test. The significance of the differences among the pairs of treatment means was evaluated by the least significance difference (LSD) test at 1% level of probability for the interpretation of results.