The recent observation of a power–law relationship, S ∝ Az, between number of taxa, S, and area, A, for microbial eukaryotes and bacteria suggests that this is one of the few generic relationships in ecology, applicable to plants, animals and microbes. However, the rate of increase in the number of species with area varies from approximately the fourth (z = 0.26) to as little as the 50th root (z = 0.0019) in microbes. This is an enormous range for which no quantitative explanation has been proffered. We show by sampling from synthetic populations that the disparity between sample and community sizes in microbial community surveys means z can be considerably underestimated and accrual of rare taxa with increasing area will not be detectable. Significant microbial taxa–area relationships will only be observed when changes in community structure within samples correlate with area. Thus, the very low z values observed recently cannot be used as the sole evidence in support of any particular community theory of community assembly. More generally, this suggests that our search for patterns and laws in the microbial world will be profoundly influenced and, potentially distorted by the sample sizes that are typical of microbial community surveys.