When people estimate quantities of real-world objects (e.g., caloric content) Brown and Siegler (1993) proposed that they rely on their metric knowledge (among other things) to do so. Metric knowledge is defined as knowledge about the statistical properties of a domain and the quantity being estimated, such as the mean, variance, as well as the form of the distribution, which is distinct from knowledge about individual objects. Multiple studies have shown exposure to numerical facts about a domain can enhance people’s metric knowledge and estimation accuracy—a phenomenon known as the seeding effect. However, existing measures of metric knowledge typically rely on item-wise deviations between true and estimated values, which fail to capture participants' understanding of the overall distribution. To address this limitation, I developed a novel task that intuitively and flexibly assesses participants' knowledge of a domain's full distribution. In an initial experiment, participants estimated the distributions of two real-world quantities (calories in food items and days until sexual maturity in mammals). They then received numerical information about a subset of items and were allowed to revise their distributions. Results revealed that on average participants initially exhibited limited knowledge of the true distributions but were able to update their understanding in a normative manner following exposure to the numerical facts. These findings provide a promising approach to studying distributional knowledge and its role in estimation accuracy.