Any communication channel will usually distort the transmitted signal. This is especially true in the case of mobile systems, where multipath propagation causes the received signal to be seriously degraded. Over the years, many techniques have been proposed to combat channel effects. Two of the most popular are linear equalisation (LE) and decision feedback equalisation (DFE). These methods offer a good compromise between performance and computational complexity. LE and DFE are implemented using finite impulse response (FIR) filters whose frequency spectrum approximates the inverse of the channel spectrum plus noise. In mobile systems, the equaliser is made adaptable in order to be able to respond to the channel variations. Adaptability is achieved using adaptive FIR filters whose coefficients are iteratively updated. In principle, an infinite number of filter coefficients would be needed to achieve perfect channel inversion. In practice, the number of taps must be finite. Simulations show that, in realistic scenarios, making the equaliser longer than a certain (undetermined) number of taps will not yield any benefit. Moreover, computation and power will be wasted. In battery powered devices, like mobile terminals, it would be desirable to have the equaliser properly dimensioned. The equaliser's optimum length strongly depends on the particular scenario, and as channel conditions vary, this optimum is likely to vary. This thesis presents novel techniques to perform equaliser length adjustment. Methods for the LE and the DFE have been developed. Simulations in many different scenarios show that the proposed schemes optimise the number of taps to be used. Moreover, these techniques are able to detect changes in the channel and re-adjust the equaliser length appropriately.