In order to monitor barley and weed development on a loamy sand soil subjected to different agronomic practices, field experiments were conducted for three growing seasons (1992–95) in a semi‐arid agrosystem in central Spain. For eight years, independent plots were managed with three crop sequences: barley (Hordeum vulgareL.)?vetch(Vicia sativaL.); barley? sunflower (Helianthus annuusL.); and a barley monoculture. In all cases, two tillage systems were implemented: no‐tillage and conventional tillage. In the years with standard rainfall (400 mm) an improvement in growth‐related cultivation variables and yield components of barley were observed in plots under barley?vetch rotation and/or conventional tillage. In drier conditions (<350 mm) the growth rate, crop yield and yield components of barley tended to improve under the no‐tillage system. Barley?vetch rotation and/or conventional tillage increased nitrate‐nitrogen (NO3‐N) content in barley plants. Similar results were found for the concentrations of nitrogen (N), phosphorus (P), potassium (K), calcium (Ca), and magnesium (Mg). In addition, the plots under crop rotation showed a weed biomass with a high concentration of N, K, and Ca in comparison with plots under monoculture. The lack of spring herbicide treatment in the no‐tillage plots led to a 3‐fold increase in weed density compared with the plots under conventional tillage.