Background and Objective <p>Achieving adequate concentrations of beta-lactam antibiotics in patients with severe infections is challenging, and therefore therapeutic drug monitoring (TDM) would be optimal for guiding beta-lactam dosing, but is often limited by availability. UV/VIS spectrophotometry can quantify drug levels in saline, and if those levels truly reflect plasma concentrations, it could enable bedside TDM. We set out to determine whether ultraviolet and visible light range (UV/VIS) spectrophotometry can accurately estimate plasma meropenem concentrations in patients with severe infections compared with liquid chromatography with tandem mass spectrometry (LC‑MS/MS).</p> Methods <p>Blood samples were collected before the first meropenem dose, or if not possible, at least 24 h after the last meropenem dose (baseline), and 2 h after administration. Total plasma concentrations were determined by LC‑MS/MS.</p> Results <p>A total of 36 patients and 59 samples were included. The correlation between absorbance and meropenem concentrations was <i>r</i>² = 0.44 in after administration samples and <i>r</i>² = 0.18 for the difference between after administration and baseline samples. In Bland–Altman plots the bias and limits of agreement were 0.00 (−29.7 to 29.7) mg/L when meropenem concentrations were estimated using the equation on the basis of the correlation between absorbances and meropenem concentrations, 1.32 (−28.5 to 31.2) mg/L when the standard curve in saline was used to estimate meropenem concentrations, and 2.6 (−23.8 to 29) mg/L when using the correlation between the difference in absorbance versus meropenem concentrations between after administration and baseline samples.</p> Conclusions <p>In patients with severe infections, UV/VIS spectrophotometric estimation of plasma meropenem concentrations currently lacks the precision to estimate meropenem measured by LC‑MS/MS.</p> Trial Registration <p>NCT04282785, 25/02/2020.</p>