Aims: To assess optimal parameters and procedures for respiratory gated radiotherapy by using an organ motion model. Methods: A parameterised model is presented that describes the duty cycle as a function of a number of parameters; the planning position, the machine gating position and the margin size. Respiratory motion is simulated by both simple sinusoidal motion and a more clinically realistic breathing model that allows the target position at exhalation to be modelled more accurately. Patient data were used to assess uncertainties in motion amplitude and phase reproducibility. The optimal gating position was evaluated based on NTCP calculations and as a compromise between a decrease in residual motion and increased delivery time. Results: The maximum duty cycle can be achieved by choosing end of exhale as the planning position and using a machine gating position shortly before the exhalation position. The exact choice of these positions depends on the acceptable dose variation and on the margin size. Planning at the exhalation position also results in a smaller residual motion within the beam on period for a selected duty cycle. Clinical respiratory data show that the exhalation positions are most reproducible and have the smallest variation in motion amplitude (2.2 mm at exhalation versus 5.9 mm at inhalation). Gating at inhalation has the advantage of the lung volume being at its largest. However, the position at which the fractional treated lung volume is smallest depends on the margin size and the individual lung volume increase. A comparison of the NTCP indicates only a minor dosimetric advantage for gating at inhalation. Conclusion: The model has the potential to indicate whether the tumour mobility justifies implementing gating for a particular patient. If gating is justifiable then the model assists the choice of the most appropriate gating parameters. The results suggest end of exhalation for gated beam delivery.