BACKGROUND: Since 2010, countries in the African meningitis belt have been introducing a new serogroup A meningococcal conjugate vaccine (MenAfriVac(®)) through mass campaigns. With the subsequent decline in meningitis due to Neisseria meningitidis serogroup A (NmA) and relative increase in meningitis due to other serogroups, mainly N. meningitidis serogroup W (NmW), the World Health Organisation (WHO) initiated a review of the incidence thresholds that guide response to meningitis epidemics in the African meningitis belt. METHODS: Meningitis surveillance data from African meningitis belt countries from 2002 to 2013 were used to construct a single NmW dataset. The performance of different weekly attack rates, used as thresholds to initiate vaccination response, on preventing further cases was estimated. The cumulative seasonal attack rate used to define an epidemic was also varied. RESULTS: Considerable variation in effect at different thresholds was observed. In predicting epidemics defined as a seasonal cumulative incidence of 100/10(5) population, an epidemic threshold of 10 cases/10(5) population/week performed well. Based on this same epidemic threshold, with a 6 week interval between crossing the epidemic threshold and population protection from a meningococcal vaccination campaign, an estimated 17 cases per event would be prevented by vaccination. Lowering the threshold increased the number of cases per event potentially prevented, as did shortening the response interval. If the interval was shortened to 4 weeks at the threshold of 10/10(5), the number of cases prevented would increase to 54 per event. CONCLUSIONS: Accelerating time to vaccination could prevent more cases per event than lowering the threshold. Once the meningitis epidemic threshold is crossed, it is of critical importance that vaccination campaigns, where appropriate, are initiated rapidly.