Invertebrate pests continue to cause substantial crop losses, while the standard methods of controlling them are often unsustainable. Capturing the services of natural pest control is one way to mitigate these losses. Such services are estimated to be worth ~$400 billion annually worldwide. Perennial habitat is increasingly recognized for the important role of providing habitat for beneficial insects and making them available to move into crops. This research focused on better capture of pest control services by investigating: the impact of early arrival and predation on pest populations in the Lockyer Valley, how pest suppression was affected by crop and non-crop habitat in the surrounding farm and landscape, the contribution of beneficial insects from on-farm refuges, and the voracity of two common predators on aphid pests. Conducting extensive field experiments on growers’ properties throughout the Lockyer Valley, and in the laboratory, we show that: – The landscape surrounding fields and farms can greatly influence the abundance and movement of beneficials into crops. – Landscapes with > 10% Lucerne provided more predators into the crop early, keeping pest populations lower for longer. – Pest populations exploded in the absence of beneficials. – Native vegetation is a source of beneficials, and may be important as a refuge when few crops are growing. – Growers can enhance beneficial populations by providing on-farm refuge. – An on-farm refuge for beneficial insects results in more beneficials accumulating in the crop. This effect is more pronounced when the landscape contribution of beneficial insects is low. – Two common predators of crop pests consume 39 and 60 aphids per day, respectively, and the more aphids available the more they eat. These findings show that early arrival of predators into crops keeps pest populations lower for longer. However, landscapes vary in the abundance of beneficial insects available to colonize. As part of an IPM program, growers can enhance pest suppression by managing an on-farm refuge. Future R&D should focus on trialling best-bet on-farm refuge options for vegetable production systems, and the development of accompanying decision-support tools to assist growers with plant selection, implementation and management.