We evaluated the predictive value of the COPD assessment test (CAT™) for exacerbation in the following six months or time to first exacerbation among COPD patients with previous exacerbations. COPD outpatients with a history of exacerbation from 19 hospitals completed the CAT questionnaire and spirometry over six months. Exacerbation events were prospectively collected using a structured questionnaire. The baseline CAT score categorised into four groups (0-9, 10-19, 20-29, and 30-40) showed strong prediction for time to first exacerbation and modest prediction for any exacerbation or moderate-severe exacerbation (AUC 0.83, 0.64, and 0.63 respectively). In multivariate analyses, the categorised CAT score independently predicted all three outcomes (p = 0.001 or p < 0.001). Compared with the lowest CAT score category, the higher categories were associated with significantly shorter time to first exacerbation and higher exacerbation risks. The corresponding adjusted median time was >24, 14, 9, and 5 weeks and the adjusted RR was 1.00, 1.30, 1.37, and 1.50 in the category of 0-9, 10-19, 20-29, and 30-40 respectively. Exacerbation history (≥2 vs. 1 event in the past year) was related to time to first exacerbation (adjusted HR 1.35; p = 0.023) and any exacerbation during the study period (adjusted RR 1.15; p = 0.016). The results of this study support the use of the CAT as a simple tool to assist in the identification of patients at increased risk of exacerbations. This could facilitate timely and cost-effective implementation of preventive interventions, and improve health resource allocation. Trial registration Clinicaltrials.gov: NCT01254032.