The appropriate fire return interval of prescribed burns is critical in many fire-driven systems. I use a mediterranean-type ecosystem – Cape fynbos, South Africa – to test guidelines for fire return intervals. Guidelines state that at least 50% of a population of the slowest-maturing obligate seeder Protea species must have flowered for three successive seasons, or that 90% of individuals from the slowest-maturing Protea species must have flowered at least once before a fire occurrence. Whether these ‘rules of thumb’ will lead to similar fire return intervals for prescribed burns is unknown. I address this by assessing two obligate seeding Protea species for length of juvenile period and age at three times flowering at 17 sites across a conservation area. For the slowest maturing species, when 50% of the fastest-maturing population had flowered three times (at 11 years), no individuals had flowered three times (happens at 14 years) in the slowest-maturing population. Using the 90% rule, this was reached at 9 years in the fastest-maturing P. neriifolia population and 11 years in the slowest-maturing. The 50% rule is more conservative than the 90% rule and is preferred. I highlight the importance of testing guidelines for fire return intervals to ensure effective biodiversity conservation.