Not all studies have empirically supported the model that predicts a positive relationship between habitat heterogeneity and biodiversity. We hypothesized that these different results stem from the methods used to assess habitat heterogeneity; many studies used variables that are somewhat correlated in nature and measure 2 different features of the environment: a) the number of structure types (habitat heterogeneity) and b) the number of structures, disregarding their types (habitat amount). We tested this hypothesis with a single experiment that assigned orthogonal values of habitat heterogeneity and habitat amount to artificial environments located on the floor of a forest remnant. We statistically controlled the number of individuals in each environment to prevent a random sample effect. We used the number of arthropod morphospecies present in the environments after 60 d as our dependent variable. The results indicate that habitat heterogeneity had no significant effect on species richness, while habitat amount showed a positive effect when the number of individuals was not controlled. Neither habitat heterogeneity nor habitat amount affected species richness when the number of individuals was controlled. We conclude that conflicting results in previous tests of the heterogeneity model could stem from conceptual and methodological problems in experimental conception. We suggest that further studies distinguish between heterogeneity and area effects, design proper controls for different effects, and consider the spatial scale of the ecological processes that influence species diversity.