A simple and easily implemented method is presented here for automatic on-board estimation of a critical parameter of the control jet (thruster) impulse transfer characteristic. This facilitates optimal control with regard to thruster lifetime, fucel consumption and pointing accuracy, despite uncertainties in the thruster performance, The new estimator facilitates automatic, in-orbit adaptation of the attitude control system to slow changes in the control jet characteristics, thereby maintaining optimal performance. The parameter estimator is intended for microprocessor implementation and assumes the use of a state estimator. Attention is restricted to a single, isolated axis. No difficulty is envisaged in developing the technique further to cater for three-axis stabilised spacecraft with significant inter-axis coupling The parameters to be determined are the intercepts on the characteristics of each control jet relating the actual jet firing time to that demanded. A mismatch in these parameters gives rise to a state estimator transient following a jet firing. This provides the stimulus to the parameter estimator, which is basically a resettable pure integrator. The intercept estimate is updated by an amount proportional to the integral of the state estimator error taken over a finite period, commencing at the beginning of the jet firing. Suitable choice of the gain of proportionality yields dead-beat parameter estimation. Besides compensating for jet firing time mismatch, the control system is shown to be capable also of compensating for uncertainties in the thrust level for short turn-on times.