Environmental filter models have been proposed as conceptual organizing frameworks for comparing and contrasting restoration practices. I evaluate two such environmental filter models, one proposed by Fattorini and Halle (2004) and the other by Hobbs and Norton (2004). These models were developed by abstracting restoration practice into what the authors viewed as the essential features restoration practitioners target for control or manipulation. In so doing, these conceptual frameworks hope to be able to transfer insights between different kinds of ecosystems. Here, I take the opposite approach: given an environmental filter model, I asked how well its filters could characterize restoration practices reported in the literature. I found that it was easier to characterize specific restoration practice using the more detailed filters described by Hobbs and Norton. I found that manipulation of biotic filters was most common in terrestrial ecosystems, whereas manipulation of abiotic filters was more common in wetland and stream ecosystems. Fattorini and Halle’s model appears most useful for evaluating the current status of degraded ecosystems compared to nondegraded ones, but Hobbs and Norton’s model is better for evaluating what particular restoration activities might be undertaken to move that system from a degraded to a nondegraded state.