Conservation tillage methods, including zero-tillage, rotary-tillage, harrow-tillage and subsoil-tillage, are widely used in Northern China because of their positive effects on soil fertility and water conservation-which provide greater economic benefits than conventional tillage. Tillage methods affect not only weed biodiversity and density but also crop growth. Reasonable control of weeds is an important guarantee for higher crop yield. However, the effects of the interactions between tillage methods and weeds on crop yields in Northern China are not clear. To investigate the effects of different soil tillage methods-with and without herbicide-on weed density, dominant weed species, and the yield of winter wheat, an experiment was conducted in 2008 to 2009 on a winter wheat-summer maize cropping system incorporating two crops a year. The wheat-summer maize cropping system is the main cropping system in the region. The experiment took place at the Agronomy Research Base of Shandong Agricultural University, as part of a long-term experiment begun in 2003. The Shannon diversity index, Shannon evenness index and a species richness index were used to investigate weed biodiversity. A split plot design was used incorporating five soil tillage methods: zero-tillage, rotary-tillage, harrow-tillage, subsoil-tillage, and conventional tillage. Seven weed species were recorded in the wheat crop: Descuminia sophia (L.) Webb ex Prantl, Capsella bursa-pastoris (L.) Medik, Avena sativa L., Convolvulus arvensis L., Cirsium setosum (Willd.) Bieb., Stellaria media (L.) Vill., and Lithospermum arvense L. Without herbicide, weed diversity under zero-tillage and subsoiling was significantly higher than under other tillage methods (LSD, P＜0.05). With herbicide, weed diversity decreased significantly. Without herbicide, the dominant weed species in the no-tillage, subsoil-tillage and conventional tillage plots were Descuminia sophia (L.) Webb ex Prantl and Capsella bursa-pastoris (L.) Medik, while in the rotary-tillage and harrow-tillage plots the main species was Descuminia sophia (L.) Webb ex Prantl. With herbicide, the dominant weed species under all tillage methods was Descuminia sophia (L.) Webb ex Prantl. Without herbicide, species richness and evenness of weed community under harrow-tillage and conventional tillage were higher than with other tillage methods. Weeds and their interaction with tillage methods significantly affected wheat yield and composition. Wheat grain yields were increased with the use of herbicide, and subsoiling tillage gave the highest grain yield of the five tillage treatments. At the heading stage, the height of weed plants was close to or even exceeded that of the wheat, which could induce light competition between weeds and wheat resulting in decreased grain yield. Without herbicide, increasing weed biodiversity decreased weed density, while the wheat yield was still lower than in plots treated with herbicide. Subsoiling with herbicide gave higher wheat yield, mainly because of the effects of tillage methods, weeds and their interactions on light conditions and other resources.