It is puzzling today to explain both the diversity and the rationale of regulators’ practice vis-à-vis network monopolies. We argue that two fundamental characteristics should be considered when defining the most appropriate regulatory tools. First, it is the bounded endowment of regulators set by governments and legislators which determines their abilities (staff, budget, administrative powers) to implement any of the regulatory tools. Ranked from the easiest to the most demanding to implement, these various tools are: a- cost plus, b- price/revenue cap, c- output or performance-based regulation, d- menu of contracts and e- yardstick competition. Second, the regulators also have to take into account that the network monopolies perform multiple tasks with heterogeneous regulatory characteristics (in terms of controllability, ex ante predictability and ex post observability). These characteristics of tasks determine what type of regulatory tool is more likely to better regulate each task. The regulatory tools then perform well only when they are implemented for tasks that are controllable and predictable enough. It is the kind of observability of these tasks which determines the best incentive tool to implement. Lastly, conclusions for the regulation of networks are derived. A workable regulation of network relies on a reasonable alignment of the regulatory tools with the regulatory characteristics of tasks and the regulators resource endowment.