A ramp metering control algorithm includes typically two distinct components: the (most important) control strategy, which makes real-time decisions on the ramp exit flow to be implemented, and the translation of this decision into specific traffic light settings according to the applied metering policy. This study focuses on the second component, providing an overview of practiced metering policies (one car per green, n cars per green, full cycle, and discrete release rates) along with a discussion of their advantages and shortcomings. Further related issues, such as traffic signals, multiple metering lanes, bypass lanes, check-in and check-out detectors, integration of both control components, and on-off switching logic, are also addressed.