Internet-of-Things (IoT) is envisioned to provide connectiv-ity to a vast number of sensing or actuating devices with limited com-putational and communication capabilities. For the organizations that manage these constrained devices, the monitoring of each device's oper-ational status and performance level as well as the accounting of their re-source usage are of great importance. However, monitoring and account-ing support is lacking in today's IoT platforms. Hence, this paper stud-ies the applicability of the Constrained Application Protocol (CoAP), a lightweight transfer protocol under development by IETF, for eciently retrieving monitoring and accounting data from constrained devices. On the infrastructure side, the developed prototype relies on using stan-dard building blocks oered by the AMAAIS project in order to collect, pre-process, distribute, and persistently store monitoring and account-ing information. Necessary on-device and infrastructure components are prototypically implemented and empirically evaluated in a realistic sim-ulation environment. Experiment results indicate that CoAP is suited for eciently transferring monitoring and accounting data, both due to a small energy footprint and a memory-wise compact implementation. Abstract. Internet-of-Things (IoT) is envisioned to provide connectivity to a vast number of sensing or actuating devices with limited computational and communication capabilities. For the organizations that manage these constrained devices, the monitoring of each device's operational status and performance level as well as the accounting of their resource usage are of great importance. However, monitoring and accounting support is lacking in today's IoT platforms. Hence, this paper studies the applicability of the Constrained Application Protocol (CoAP), a lightweight transfer protocol under development by IETF, for efficiently retrieving monitoring and accounting data from constrained devices. On the infrastructure side, the developed prototype relies on using standard building blocks offered by the AMAAIS project in order to collect, pre-process, distribute, and persistently store monitoring and accounting information. Necessary on-device and infrastructure components are prototypically implemented and empirically evaluated in a realistic simulation environment. Experiment results indicate that CoAP is suited for efficiently transferring monitoring and accounting data, both due to a small energy footprint and a memory-wise compact implementation.