It is shown that the performance of a given active filter configuration is dependent upon the design procedure employed. Consequently, design procedures may be selected to optimize some aspect of the active filter's performance. This paper develops design procedures which minimize the effect of gain bandwidth (GB) upon a given active filter configuration. The root sensitivity with respect to the GB can be decreased by a factor of two or more depending upon the realization. The techniques presented are applicable to most single and multiple amplifier, second-order active filter realizations.