In order to increase CRs (Cognitive Radios) operation efficiency, there has been an interest in increasing awareness level about spectrum utilisation. In this respect, this paper proposes a new fittingness factor concept that captures the suitability of spectral resources exhibiting time-varying characteristics to support a set of heterogeneous CR applications. Different fittingness factor functions to track unknown variations of interference levels are formulated and analysed. First, the dependency with traffic load is studied and second, the impact over the spectrum selection decision-making process in a multiservice CR context is evaluated. Results show that, even with a simple greedy approach, the fittingness factor concept can result in an efficient matching of spectral resources to the requirements of CR applications, thus resulting in significant reduction in the user dissatisfaction probability.