To compare the effects of different weed management methods on seed bank of weeds grown in sugar beet (Beta vulgaris L.), a field experiment was conducted at the Agricultural Research Station, Ferdowsi University of Mashhad, Iran, during 2006, 2007 and 2008. The experiment designed as a complete randomized block with three replications. The treatments including: disk+ fhenmedifham, disk+ fhenmedifham+ rotation, disk+ cultivation, disk+ cultivation+ rotation, metamitron+ fhenmedifham, metamitron+ fhenmedifham+ rotation, metamitron+cultivation, metamitron+ cultivation+ rotation, covercrop+ fhenmedifham, covercrop+ cultivation, weeding , weeding+ fhenmedifham and control. Soil samples were taken (in the base of clustering method) in two depths of 0-10 cm and 10-20 cm, at two stages, early season and late season. Results showed that at the first experiment highest and lowest seed bank size obtained in control plots (383 seeds per kg of soil) and weeding+fhenmedifham (52 seeds per kg soil). At the second year maximum and minimum seed bank size were observed in control and weeding treatment with 1080 and 319 seeds/kg soil, respectively. In two years, seed bank size in 0-10 cm depth was more than 10-20 cm depth. In the first and second years seed bank size at late season compare to early season was 2.2 and 1.7 times, respectively. In two years, significant coefficients were observed in seed bank size at early season and late season (R2=0.76 and R2=0.65). There were significant correlations between seed bank size in the first and second year (R2=0.71) and between seed bank size and frequency of seedling (R2=0.77).