Decentralised manufacture (DM) of personalised medicines via 3D printing (3DP) is now viable in the United Kingdom under new legislation with regulatory agencies across the globe working to follow suit. Quality control (QC) and ensuring dose accuracy for small-batch production of extended-release (ER) printlets (3DP tablets) remains a challenge. We demonstrate, for the first time, a non-destructive, volumetric method for drug quantification using Transmission Raman spectroscopy (TRS) combined with partial least squares regression (PLSR), suitable for high-throughput QC of individual ER printlets in DM settings. A simple theophylline-based pharma-ink was developed for direct powder extrusion (DPE) 3DP of cylindrical ER printlets in 3 sizes, maintaining a constant surface area-to-volume (SA/V) ratio to ensure comparable release profiles. In vitro dissolution confirmed equivalent ER profiles across all sizes, meeting pharmacopeial requirements. A PLSR model calibrated with only the smallest and largest printlet TRS spectra accurately predicted drug content in all 3 printlet sizes (R² = 0.9948, RMSEP = 0.5611% w/w) with no statistical differences compared to the reference method. These findings establish TRS as a rapid, non-destructive drug content QC tool for personalised, dose-flexible 3DP medicines and support its implementation in decentralised pharmaceutical manufacturing workflows.