Spatial distribution of Sorghum halepense (L.) Pers. populations was assessed in tomato cropping fields in a total of 11 commercial fields (93 ha). Weed infestation was visually assessed from the cabin of a tractor after harvesting, using a three category ranking, ‘high’, ‘low’, and ‘no presence’, through infestation maps. Crop management factors as well as intrinsic parameters of patches were collected and calculated. The proportion of the field infested with low and high S. halepense densities, patch anisotropy, the effect of field borders and field topography were studied. On average, 5 and 3% of the surveyed area was infested with high and low densities, respectively. The majority of patches were of small size and most of the infested area was concentrated in a few large patches with irregular shape. Small patches, those with less than 50 m2 , represented 70% of the total number of detected patches. However, they only accounted for the 3% of infested area. Tillage operations showed a great influence on patch shape, producing patches twice longer in the direction of tillage than perpendicular to tillage. This result revealed the influence of human operations in S. halepense spreading. The effect of edges also had a great influence in patch expansion. Patches in contact with a field border were almost five times longer than their width in the direction of tillage. Also, the effect of borders stimulated the infestation. Areas closer to the borders had a higher risk of S. halepense infestation than zones in the center of the fields. In addition, patches tended to increase complexity the bigger they became, with a progressive shrinkage in the ratio area/perimeter2 . The influence of location within the field revealed that higher levels of infestation were found on the lowest and closest areas to riverbeds, in areas with flooding risk. Characterizing the location of S. halepense patches after harvesting offers a precise and cheap method for the construction of weed maps, which can be used for sitespecific treatments and description of weed spatial biology. This research was funded by the Foundation Alfonso Martín Escudero. Peer reviewed