Comparisons are powerful tools for making sense of worlds. But comparisons do not merely identify inherent or pre-existing similarities and differences; instead, they participate in making the very worlds they aim to describe. This dissertation probes how comparative practices shape the formation of multispecies landscapes. I show that the specificities of how people make comparisons and what kinds of comparisons they make are a powerful but often overlooked part of the production of human-nonhuman arrangements, as well as knowledges about them. By focusing on salmon in Hokkaido, Japan, I demonstrate that neither the island's watershed ecologies nor its fish population structures can be understood without attention to practices of comparison-making. Since the mid-19th century, natural resources management in northern Japan has been profoundly shaped by how people both within and beyond Japan have compared Hokkaido's landscapes and fish to those in other parts of the world. After the Meiji Restoration of 1868, Japanese officials sought to "develop" Hokkaido's lands and waters and make them legibly "modern" to Euro-American audiences. They did so by importing "Western" crops and livestock, promoting large-scale commercial agriculture, and constructing salmon hatcheries. Since that time, the ways that Japanese fishermen, scientists, government officials, and indigenous peoples have compared Hokkaido's salmon and salmon-bearing watersheds to others around the globe have dramatically affected the region's approaches to fisheries management, as well as its salmon. Drawing on ethnographic, archival, and fisheries science research in Hokkaido, the U.S. Pacific Northwest, and Chile, my dissertation demonstrates how particular practices of comparison have created cross-border movements - such as transplants of salmon eggs, exchanges of currency, transfers of scientific technology, and exports of processed fish products - that have shaped the course of Hokkaido's development, the genes of its fish, and the identities of its people. Comparison, I show, is not just a "meta" act of analysis, but also an everyday practice that alters both human and nonhuman bodies and relations. When a Japanese consumer compares kinds of salmon at a supermarket and decides to purchase the fish labeled "wild," or when a Chilean biologist compares the temperature of a Patagonian river to one in Hokkaido, determining that it might be possible to transplant fish from one side of the Pacific to the other, such decisions fundamentally reconfigure human and salmon lives. By tracing such far-reaching comparisons, this dissertation attempts to open up the practices and geographies through which we know landscapes.