This dataset includes the pore pattern measurements of calcareous benthic foraminifera to calibrate a proxy for reconstructing bottom water dissolved oxygen (BWDO) levels in the Southeast Pacific (SEP), complemented by environmental and biogeochemical data. Surface sediment samples collected between 1962 and 2014 span depths of 24–3,252 meters and latitudes from 12°S to 44°S off the coasts of Chile and Peru. These samples were obtained using various devices, including a multicorer (Sonne 156, Sonne 211, Meteor 92), a gravity core (Sonne 161), a Petersen grab (USNS Eltanin), and a box corer (BIAC072014). Specimens of six benthic foraminifera species (Cibicidoides wuellerstorfi, C. lobatulus, C. ungerianus, Planulina ariminensis, P. ornata, and P. limbata) were collected from the sediment samples. The specimens, dated recent (Holocene to modern), were imaged using Scanning Electron Microscopy (SEM) between March and December 2022. Pore pattern measurements were conducted using ImageJ software between March 2022 and March 2023, including porosity, pore density, and pore size. This study aims to validate benthic foraminifera pore patterns as proxies for reconstructing past BWDO levels. These are crucial for understanding current and future long-term ocean oxygenation trends due to climate change. This dataset provides a region-specific calibration for BWDO proxies, offering detailed insights into geological ocean conditions with an error margin of ±5–20 µmol/kg. To complement the benthic foraminifera morphology data, environmental and biogeochemical parameters, including bottom water temperature (°C), pH (25°C NBS scale), nitrate concentration, and dissolved oxygen (µmol kg-1), were estimated using a curated hydrological dataset (Reyes-Macaya et al., 2024 https://doi.org/10.5281/zenodo.13362423). This dataset includes CTDO profiles and discrete bottle measurements collected over the past 20 years from the Equatorial and Southeast Pacific, covering depths from the surface to 6,000 meters. Data interpolation was performed using Ocean Data View software (Schlitzer, 2023 https://odv.awi.de), employing DIVA interpolation and Weighted Average Interpolation. Site-specific values for each sediment sample were extracted using a custom Python script (BOT; Reyes-Macaya et al., 2024 https://doi.org/10.5281/zenodo.13362423), identifying the closest hydrological match regarding latitude, longitude, and depth. Net primary productivity (NPP) was estimated using the Carbon-based Production Model (CbPM; Westberry et al., 2008 https://doi.org/10.1029/2007GB003078), which uses SeaWiFS-derived variables such as chlorophyll concentration, nitrocline depth, mixed layer depth, and light availability.