Tissue cultures of a medicinal plant, Plantago lanceolata L. were screened for phenylethanoid glycosides (PGs) and other natural products (NPs) with LC-ESI-MS(3). The effects of N source concentration and NH4(+)/NO3(-) ratio were evaluated in a full-factorial (FF) experiment. N concentrations of 10, 20, 40 and 60mM, and NH4(+)/NO3(-) ratios of 0, 0.11, 0.20 and 0.33 (ratio of NH4(+) in total N source) were tested. Several peaks could be identified as PGs, of which, 16 could be putatively identified from the MS/MS/MS spectra. N source concentration and NH4(+)/NO3(-) ratio had significant effects on the metabolome, their effects on individual PGs were different despite these metabolites were of the same biosynthethic class. Chief PGs were plantamajoside and acteoside (verbascoside), their highest concentrations were 3.54±0.83% and 1.30±0.40% of dry weight, on media 10(0.33) and 40(0.33), respectively. NH4(+)/NO3(-) ratio and N source concentration effects were examined on a set of 89 NPs. For most NPs, high increases in abundance were observed compared to Murashige-Skoog medium. Abundances of 42 and 10 NPs were significantly influenced by the N source concentration and the NH4(+)/NO3(-) ratio, respectively. Optimal media for production of different NP clusters were 10(0), 10(0.11) and 40(0.33). Interaction was observed between NH4(+)/NO3(-) ratio and N source concentration for many NPs. It was shown in simulated experiments, that one-factor at a time (OFAT) experimental designs lead to sub-optimal media compositions for production of many NPs, and alternative experimental designs (e.g. FF) should be preferred when optimizing medium N source for optimal yield of NPs. If using OFAT, the N source concentration is to be optimized first, followed by NH4(+)/NO3(-) ratio, as this reduces the likeliness of suboptimal yield results.