Attenuated total reflectance-infrared spectroscopy (ATR-IR) coupled with partial least squares regression (PLSR) was evaluated as a rapid, label free and cost-effective tool to quantify water content in extracts obtained from spirulina wet biomass using a glucose glycerol natural deep eutectic solvent (NADES). NADESs are green, renewable and biodegradable solvents with unique properties outcompeting existing organic solvents, for instance, for plant or biomass extraction. The properties of NADESs depend critically on their water concentration, and therefore, it is essential to develop methods to monitor it, to ensure optimal extraction efficiency and experimental repeatability to achieve a better standardization of extraction protocols. First, Karl Fischer titration was performed on a set of 20 NADES extracts in order to obtain reference water concentrations. Secondly, ATR-IR spectra were collected and subjected to datamining to construct PLSR predictive models. An R2 value of 0.9996, a mean root mean square error of cross validation of 0.136% w/w and a root mean square error of prediction of 0.130% w/w highlight the feasibility and reliability to perform quantitative analysis using ATR-IR. Moreover, the mean relative error percentage achieved, ∼0.5%, confirms the high accuracy of water concentration determination in NADES extracts. This work demonstrates that powerful alternatives are available to provide more environmentally responsible analytical protocols. ATR-IR spectroscopy applied to NADES extracts does not require any sample preparation, reagents or solvents and has minimal requirements for single use consumables. The technique is consistent with current concerns to develop greener chemistry, especially in the field of extraction of natural compounds from plants which currently represents a major focus of interest in both research and industry.