Relative abundance indices based on catch and effort data can become biased unless consideration is given to the spatial dynamics of the fishery such as changes in either the spatial distribution of fishing effort or the range of the stock over time. The construction of such indices therefore needs to take into account features of the fishery itself. In this paper, a general framework is presented for developing more appropriate abundance indices based on fishery catch and effort data. In developing this framework, it adopts the approach of (i) developing a range of hypotheses which encompass the uncertainties in the spatial–temporal dynamics of the stock and the fishing effort, (ii) identifying the hypotheses underlying the different CPUE series, and (iii) evaluating the available information relative to these hypotheses as the basis for evaluating CPUE indices. Observations from the fishery for southern bluefin tuna (Thunnus maccoyii) are used to illustrate various hypotheses about the nature of the fishery which can be used to construct indices of stock abundance while a simple simulation framework is used to explore the implications of some of these hypotheses on the accuracy of such indices.