1. The nature and resources supplied by different components of habitats influence species, creating variability from place to place within a habitat. 2. Experiments were done to investigate the effects of altering components of habitats on the variability of assemblages of numerous species of intertidal gastropods. 3. Artificial habitats with three levels of structure, combining different types of turf (i.e. different densities and height of fronds) were sampled 8 weeks after deployment in the intertidal. They were rapidly colonized by up to 66 species of gastropods. 4. Independently of the types of turf combined to form different habitats, there were differences in assemblages where there was more than one type of component present. Multivariate dissimilarities among units making up each habitat were also greater where there were more than one type of unit, but there was no such difference in the variance of numbers of species per unit. 5. Altering the relative abundances of different types of components made little change to the assemblages, nor their multivariate variability among units of habitat and the variance in number of species per unit in each habitat. 6. Differences in assemblages due to the different structure of habitat are complex to interpret and simple characterizations of structure of habitat are inadequate. Comparing different habitats requires appropriate experimentation to ensure that variability within habitats is also investigated.