During the past two decades, the comet-based in vitro DNA repair assay has been used regularly to measure base excision repair (BER)-related DNA incision activity. Most studies focus on the assessment of BER in human lymphocytes or cultured cells by estimating the activity of a cell extract on substrate DNA containing specific lesions such as 8-oxoguanine. However, for many 'real-life' studies, it would be preferable to measure BER in the tissues of interest instead of using in vitro models or surrogate 'tissues' such as lymphocytes. Various attempts have been made to use the comet-based repair assay for BER with extracts from rodent tissues, but high non-specific nuclease activity in such tissues were a significant impediment to robust estimates of BER. Our aim in this study was to optimise the in vitro repair assay for BER for use with rodent tissues using extracts from liver and brain from C57/BL mice. Because the DNA incision activity of an extract is dependent on its protein concentration, the first optimisation step in preventing interference by non-specific nuclease activity was to determine the protein concentration at which there is a maximal difference between the total and non-specific damage recognition. This protein concentration was 5 mg/ml for mouse liver extracts and 1 mg/ml for brain extracts. Next, we tested addition of proteinase inhibitors during the preparation of the tissue extracts, but this did not improve the sensitivity of the assay. However, addition of 1.5 μM aphidicolin to the tissue extracts improved the detection of DNA repair incision activity by reducing non-specific nuclease activity and possibly by blocking residual DNA polymerase activity. Finally, the assay was tested on tissue samples from an ageing mouse colony and in mice undergoing dietary restriction and proved capable of detecting significant inter-animal differences and nutritional effects on BER-related DNA incision activity.