<p>The pulp and paper industry is an active player in the European bio-economy producing wood-based, biodegradable material. The recovery of chemicals spent in the cooking step separating cellulose from the lignocellulosic biomass is a key step towards circular processing. However, seasonally varying input streams, complex chemical composition, and intricate interaction between gas and liquid phase result in precipitation of insoluble salts. As a result, clogged pipes lead to unscheduled downtimes as well as loss in valuable chemicals, challenging the economic feasibility and ecological footprint of the process. Here, Raman spectroscopy is used as a non-destructive, in situ process monitoring tool, to help deepen the understanding of the chemical interplay occurring in the chemical recovery process. In combination with multivariate regression models based on several hundred reference spectra, the spectral fingerprint is translated into critical process-relevant parameters. A workflow to set up such a multivariate model in industrial environment is shown and successful test-wise online implementation at the plant in Gratkorn, Austria, is demonstrated. Continuous monitoring allows tight process control of the interaction between gas and liquid phase. Hence, effects due to e.g. seasonally varying input streams resulting in varying chemical composition may be corrected or avoided by timely countermeasures, overall reducing unscheduled downtimes and loss of valuable chemicals.</p> Graphical Abstract <p></p>