Mangrove sediments are valuable archives of paleoenvironmental and relative sea-level changes. The most widely applied method to obtaining chronologies of past changes in mangrove sediments is radiocarbon dating, because mangroves produce large amounts of organic matter in situ. However, there are many challenges to obtaining reliable radiocarbon chronologies because bioturbation processes from roots and crabs can rework mangrove sediments, resulting in ages that are not in stratigraphic order. Previous studies have suggested that methods that isolate specific sediment size fractions may yield ages closer to the age of the paleo depositional surface by removing younger carbon contamination from fine roots. This study examines which sample types are more likely to yield reliable radiocarbon ages using shallow cores from a mangrove environment on Mahé, Seychelles, in the Indian Ocean. We compare radiocarbon ages from bulk sediment, sieved organic concentrates and above-ground macrofossils collected from the same stratigraphic depths. Bulk sediment and organic concentrate ages are comparable, which suggests that methods that separate out different size fractions do not sample different carbon sources in Seychelles mangrove cores. Identifiable above-ground macrofossils are rare in Seychelles mangrove cores, but yield older radiocarbon ages than comparable bulk sediment or organic concentrate ages. We suggest that in Seychelles, limited accommodation space over the late-Holocene, determined by relatively stable relative sea levels, has resulted in poor preservation of above-ground macrofossils for radiocarbon dating due to low rates of burial and sediment accretion. Low accretion rates have likely resulted in a mangrove sediment sequence that is highly bioturbated and degraded, meaning both bulk sediment and organic concentrate samples are impacted by contamination from younger roots. We argue that the availability of accommodation space and sediment composition controls the reliability of mangrove radiocarbon chronologies, which has implications for sample choice and site selection.