Anxiety and depression are common mental health problems in later life. Since worry and rumination are thought to underpin the respective primary cognitive processes in anxiety and depression, we developed a measure to distinguish worry from rumination in later life. The Ruminative Response Scale was adapted to include items that characterise the cognitive features of worry. We examined its properties using 92 clinical and non-clinical participants, aged over 65. Factor analysis demonstrated a three-factor structure: brooding, reflection and worry with internal consistencies of alpha = 0.72, alpha = 0.67 and alpha = 0.55 respectively. We found no evidence for concurrent validity of these factors using the Penn State Worry Questionnaire. Modest but significant associations between reflection and brooding (r = 0.36) and reflection and worry (r = 0.2) were found. Brooding and worry sub-scales remained unrelated. We suggest that it is possible to distinguish worry from rumination in older people and that differentiating between their key underlying characteristics in the assessment of mood problems may enhance the targeting and evaluation of cognitive-behavioural therapy for anxiety and depression in later life. Future research with a substantial clinical sample is needed to explore the underlying dimensions and correlates of worry in later life.