A considerable body of evidence has demonstrated the positive impact of the diversity of plant species on biological control of pests at the field scale. However, very few studies have assessed the effect of crop diversity on natural enemy performance and pest control. In order to test our hypothesis that crop diversity could increase natural enemy development and performance, we examined the mechanisms underlying the effect of two types of crop succession, i.e., multiple-crop succession (tomato, squash and soybean) and mono-crop succession (each crop alone), on population dynamics, predation capacity and spillover of Nesidiocoris tenuis Reuter (Hemiptera Miridae) in a greenhouse experiment. We found that (1) the polyculture supported lower population growth of N. tenuis and lower predation rates of Ephestia kuehniella Zeller (Lepidoptera Pyralidae) eggs compared to tomato and squash monocultures, but that correspondingly (2) the predator performed better on the polyculture than on the soybean monoculture. These results revealed that crop identity within the succession is a major factor in determining population dynamics and biological control. We found that compared to soybean monoculture, the presence of soybean Glycine max L. (Fabales Fabaceae) in the polyculture treatment reduced the population dynamics of the mirid predator but increased biological control. This result suggests that non-host crops in a polycultural succession could benefit from the natural enemy populations that were increased by other suitable crops in the succession.