Human exploitation of intertidal marine invertebrates is known to alter benthic community structure. This study describes the impact that harvesting by women and children has on the intertidal community structure of the mudflats of the Saco on Inhaca Island, Mozambique, by comparing the benthic communities of exploited and unexploited areas. Sampling was done using two different methods: a standard core sampling method with sieving over a 1 mm mesh, and collection by the women of all animals within a 10x10 m quadrat. A number of predictions were tested on the basis of previous studies. Species richness analyses were based on multivariate non-parametric tests. The women collected in total 64 different species per tide period, their mean catch amounting to 30.5 g AFDW (Ash-Free Dry Weight) per person, 69 f which comprised predators. The crab Portunus pelagica was the dominant species in weight. Average harvest was estimated at 0.12 g AFDW m-2 y-1. The benthic community comprised 117 different species, totalling 2200 organisms m-2 or 6.0 g AFDW m-2. No impact of the exploitation could be detected for any of the following macrobenthos parameters: biomass, abundance, density or percentage of target species, predator:prey ratio, target species size, species richness, species-rank abundance, or biomass/abundance ratio. No significant correlation was found between the benthos variables obtained with the core samples and the human predation pressure. The low biomass figures for predators obtained in the Saco could be linked with the human exploitation there. The samples collected in the 10x10 m quadrats comprised <0.1 f the numerical abundance and <5 f the biomass of the core samples. Species composition was also different by this method. Nevertheless, people preferred the areas with the highest abundance, biomass and species richness as judged from these 10x10 m samples. These samples apparently represented better the substrate characteristics from a human-predation perspective, and measured target species availability. The absence of any large exploitation impact is attributed to: the low level of harvest in relation to availability, the high production/biomass ratio of tropical benthos, and the mobility of the crabs. The human impact is also relatively low because of the specific characteristics of the sand-and mudflats, in which animals targeted can hide and escape predation, in contrast to rocky substrates (on which the majority of previous studies have concentrated). No keystone species could be found, and no cascading effects were apparent through the exploitation of predatory crabs.