This presentation addresses a criticism of corpus-based approaches to critical discourse studies, namely that the CL analysis does not take account of the relevant context, and shows how a preliminary corpus-based analysis can pinpoint salient contextual elements, which can inform both the CL and CDA analyses. The discussion also focuses on the importance of the statistical identification of diachronic trends (in particular, frequency peaks and troughs), and the need for high granularity in diachronic corpora. The paper aims to contribute to the synergy between CL and CDA approaches, and between qualitative and quantitative techniques in general. The presentation uses a recently completed ESRC-funded project as a case study, The Representation of Islam in the UK Press, which used a diachronic corpus of topic-specific articles. Periods of increased frequency in the number of corpus articles were identified through a statistical analysis. These frequency peaks indicate short periods (months) of significantly increased reporting on the topic/entities in focus. These periods can then be matched with events which are expected to have triggered the increased interest. This identification has a dual benefit: a) it suggests the contextual background against which the results of the corpus analysis can be interpreted; b) it provides a reliable guide to the corpus texts that can be usefully downsampled for close (qualitative) critical discourse analysis.