The objective of this air traffic management study was to analyse the trade-off between time spacing accuracy and corresponding control effort in a potential future application of airborne separation assistance systems (ASAS). The ASAS application airborne spacing sequencing and merging was simulated in fast-time. Lead aircraft speed profiles were generated using complete descent profiles from real-time experiments. For validation purposes, three metrics were derived from real-time experiments: time spacing error (accuracy), frequency of speed adjustments (control activity), and cumulative airspeed variations (control cost). Four experimental parameters were varied: automatic and manual speed control, spacing dead-zone, guidance law dynamics time constant and initial time spacing error. A trade-off between the metrics was found for a sequence of two aircraft by comparing their variation with the experimental parameters. Corresponding 'minimum' performance requirements for the metrics are proposed: (i) time spacing error-mean less than 1.5s with 0.5 to 85% of the values between-4 and +4s (automatic mode), and mean less than 2.5s with 0.5 to 85% of the values between-6 and +6s (manual mode), (ii) frequency of speed adjustments-mean less than 1 action per minute (manual mode) and (iii) cumulative airspeed variations-mean less than 10 knots (automatic and manual modes). These requirements form a basis for investigating sequences longer than two aircraft where chain propagation effects may lead to additional constraints. Nomenclature