Parcels are usually packed in single-use packaging that the recipient discards after delivery. This is not only wasteful from a resource utilization perspective, but also requires the resulting waste to be collected and centrally disposed of (or recycled), generating even more environmental impact. Reusable packaging, on the other hand, introduces additional complexity into the planning of a logistics system, as empty packages must somehow be collected from previous recipients and reintroduced into the system. We propose a new optimization problem that handles the planning of vehicle delivery routes that also collect empty packaging and deliver it to prospective senders. The goal is to utilize the available transport capacity of these vehicles, as their cargo hold becomes progressively emptier during the day as the parcels that they started with are delivered to their recipients. We describe a heuristic optimization algorithm, consisting of a construction heuristic and a local search procedure, that can be used to compute such integrated delivery routes. We evaluate this algorithm on a set of benchmark instances and compare the resulting solutions to a scenario where parcels are still delivered in single-use packaging.