Light emitting diode (LED) lamps are increasingly being studied in cultivation of horticultural, ornamental and medicinal plants as means to increase yield, quality, stress resistance, and bioactive compounds content. Enhancing the production of metabolites for medicinal or pharmaceutical use by regulating LED intensity and spectra is a challenging subject, where promising results have been achieved. Nevertheless, some species have been poorly investigated, despite their interest as a source of medicinally active substances, with particular reference to LED effects at the plant cultivation level. This study evaluates the effects of supplementary top-light LED treatments on Taxus baccata, one of the main sources of taxane precursors. Blue, red and mixed red-and-blue spectra were tested at 100 μM m-2 s-1. Moreover, 50 and 150 μM m-2 s-1 intensities were tested for the mixed spectrum. All treatments were set for 14 hours a day and were tested against natural light as control treatment, in a controlled environment, from 19 August to 9 December 2019, this latter date representing 112 days after treatment (DAT) began. A smart monitoring and control system powered by environmental and proximal sensors was implemented to assure homogeneity of temperature, humidity, and base natural light for all the treatments. It resulted in negligible deviations from expected values and reliable exclusion of confusing factors. Biometric measurements and 1H-NMR based metabolomic analysis were performed to investigate growth and phytochemical profile throughout the trial. One-way ANOVA showed that supplemental LED lighting increased plant height and number of sprouts. Considering the mixed red-and-blue spectrum, plant height increased almost proportionally from control to 100 μM m-2 s-1 (+20% at 112 DAT), with no further increase at higher intensity. The number of sprouts was strongly enhanced by LED treatments only in the early phase (48.9 vs. 7.5 sprouts in the averaged 50, 100 and 150 μM m-2 s-1 vs. the control at 28 DAT), with no differences related to intensity in the very early stage, and more persisting effects (up to 56 DAT) for higher intensities. After the very early growth stages (28 DAT), plant vigor showed a modest although significant increase over time compared to the control, with no differences related to light intensity (0.81 vs. 0.74 of NDVI in the averaged 50, 100 and 150 μM m-2 s-1 vs. the control, across 56, 84 and 112 DAT). The different spectra tested at 100 μM m-2 s-1 showed no significant differences in growth parameters, except for a slight beneficial influence of blue (alone or with red) compared to only red for sprouting. According to the metabolomic analysis, treated plants at 28 DAT were characterized by the highest content of sucrose and aromatic compounds. Signals of a putative taxane were detected in the 1H NMR profiles of plants, which were compared to the spectrum of baccatin III standard. However, the intensity of these spectral signals was not affected by the treatment, while they increased only slightly during time. Light at 150 μM m-2 s-1 induced the strongest variation in the metabolome. Conversely, light composition did not induce significant differences in the metabolome.