Chenopodium album L. is a major weed in spring-planted crops in the temperate regions of the world. Since 2000, farmers have reported an unsatisfactory control of this weed in sugar beet fields in Belgium, France and The Netherlands. Frequently, the surviving C. album plants are resistant to metamitron, a key herbicide in this crop. Metamitron resistance in C. album is caused by a Ser264 to Gly mutation in the psbA gene on the chloroplast genome, which prevents binding of metamitron to its target site. This mutation causes also resistance to other herbicides with a similar mode of action, like metribuzin -applied in potato- and atrazine in particular. Atrazine has been applied very frequently in maize in the 1970s and the 1980s, but is now banned in Europe due to environmental reasons. The persistent use of atrazine in maize confronted Belgian and other European farmers in the early 1980s with atrazine resistant C. album with the same Ser264 to Gly mutation. The problems with atrazine resistant C. album disappeared when other herbicides were applied in maize. Unfortunately, this is not the case for metamitron resistant C. album in sugar beet, because no replacement herbicide is readily available. The history of atrazine use in maize brought up a question concerning the origin of the current metamitron resistant C. album populations. Have these populations been selected locally by regular use of metamitron in sugar beet or did the selection occur earlier by atrazine use when maize was grown in the same fields? This would have serious implications regarding the reversibility of herbicide resistance. Therefore, soil samples were collected on 16 fields with different histories: five fields with an organic management over 25 years, two fields with a history of atrazine resistant C. album, five fields with metamitron resistant C. album in sugar beet and four fields which were under permanent grassland for 10 years, preceded by a regular rotation in which sugar beet was a key crop. The seeds of C. album were extracted from the soil and germinated on a germination table. Germinated seeds were allowed to grow in a growth chamber. Metamitron resistance was determined by a chlorophyll fluorescence test and leaf material was sampled for AFLP-analysis. For all fields, estimations were made of the size of the seed bank (i.e. an indirect estimate of population size), the frequency of resistant plants and the genetic diversity of resistant and susceptible populations. The results indicate that herbicide-resistant C. album populations are persistent and maintain their adaptive capacity, challenging future management of metamitron resistant C. album.