Despite earlier efforts to understand the role played by grapsid crabs inmangroves, their importance in the structuring and functioning of suchsystems is fully appreciated, particularly with regard to small-scalestudies. The present study provides some new data on the interaction between mangroves and crabs, namely the link between the distribution of particular mangrove tree species and the distribution of certain crab species at the assemblage level. Floristicand faunistic relevés were made in Gazi Bay (2 sites) and Mida Creek(3 sites), 140 km apart on the Kenyan coast, along five transects in aseries of quadrats covering the width of the mangrove belts. Zonation ofboth mangrove vegetation and brachyuran fauna was described and heightabove datum and distance to the mainland (limit of non-flooded area)measured. The relationship between the presence and the absence of crabs andtrees was analysed using detrended correspondence analysis.Summarized, the mangrove tree zonation pattern contains four assemblageswith a particular dominant species: first a landward Avicennia marinazone, followed by a mixed zone with Ceriops tagal, Rhizophoramucronata and Bruguiera gymnorrhiza, and finally a R.mucronata zone and a Sonneratia alba zone, both of which can mixwith seaward A. marina. Ordination results show that the distributionof Neosarmatium meinerti and Sesarma ortmanni is linked to thelandward A. marina zone, that of Neosarmatium smithii, Sesarma guttatum and Sesarma leptosoma corresponds to the R. mucronata zone and that of Metopograpsus thukuhar and Sesarma elongatum to the seaward A. marina and S. albazone. There appears to be one major underlying factor in the zonation of bothcrabs and trees, with most likely a complex multiple causality. In certaincases the association between crabs and trees are causal, whereas in othercases it can be the result of an independent restriction to the same zonesby a common cause.