Continuous monitoring for cyanobacteria blooms in small, inland water bodies via in-situ sampling and analysis can be challenging not only due to the number and locations of water bodies to cover, but also due to the dynamic nature of algal growth and toxin production. Detection targets vary with cyanobacteria strains as well as physical, chemical, and biological factors. Ground monitoring also lacks consistency as sampling methods, frequency, and analytical techniques vary from region to region. However, remote sensing allows systematic data collection over a large area to identify regions with potential harmful algal growth. We introduce the Cyanobacteria Aggregated Manual Labels (CAML), a large dataset of in-situ cyanobacteria measurements for investigations of cyanobacteria detection and severity classification in inland water bodies across the United States. Relevant satellite imagery from publicly available endpoints are applicable to use when applying the CAML dataset to models. The dataset labels ground measurements of cyanobacteria cell counts at 23,570 points in U.S. inland water bodies over 2013 2021. Algorithms trained on this data could be used to estimate cyanobacteria cell counts in water bodies for timely water quality and public health interventions and to gain an understanding of environmental and anthropogenic factors associated with cyanobacteria incidence and proliferation. Data is provided in a comma-separated values (CSV) format.