Crop diversification proved to improve a number of ecosystem services, including the regulation of pest, weed and disease pressure, and therefore to decrease the need for pesticides to prevent yield losses. However, the effects from crop diversification on pest control has yet to be studied quantitatively to explore a gradient in the level of diversification. Similarly, the impacts on pesticide reduction were inferential and mostly supported by evidences from landscape scale or a few long-term experiments, but the quantitative relationship between cropping system diversity and pesticide use has been rarely addressed. Furthermore, the process of cropping system diversification can both provide pest regulation effects due to the intrinsic crop nature, and the changed diversity of crops grown, which in turn affect pesticide needs. These two effects add up to a net effect on cropping system scale which is hard to be differentiated through experimental design or modeling approach. We employed the DEPHY network database describing 1528 cropping systems and 79 cash crops to disentangle and quantify the two complementary effects on pesticide use at the cropping system level. Our results suggest that the relative effect sizes from crop species identity and crop diversity, approached by crop number per cropping system, is around 37% and 5% respectively. Excluding the effect of the nature of crops, adding one supplementary crop in the cropping system decreases pesticide use by 0.06 TFI, on average. Further studies on crop species traits effects on divided sub-group of herbicide, insecticide and fungicide are required to shed light on the effects brought by crop species identity.