Weeds are a big problem in agriculture and forestry, and herbicides are the main tools to control them. Phosphinotricin (ammonium glufosinate, PPT) is one of the most effective non-selective herbicides, to which weeds hardly gain resistance, but the reasons for its effect and toxicity to plants are still unclear, and especially, it is little studied in trees, including transgenic ones. We studied the physiological responses of downy birch (Betula pubescens Ehrh.) containing the herbicide resistance bar gene or the cytosol glutamine synthetase GS1 gene (the target enzyme of the herbicide) to PPT-based Basta herbicide treatment in various doses under open-air conditions during two years. Birch saplings with the bar gene were resistant to a double field dose (10 L/ha), but the expression of the GS1 gene only slightly increased resistance compared to the control. Herbicide treatment increased the ammonium level in leaf tissue by 3–8 times, but this, apparently, was not the main cause of plant death. Among leaf pigments, chlorophyll B was the most resistant to PPT, and carotenoids were the most sensitive. Responses of birch trees with the GS1 gene (accumulation of ammonium, pigment content, and dehydration) during treatment with a low dose of herbicide were less pronounced than in control plants. One-year-old control and transgenic plants with the GS gene died after 2.5 L/ha treatment, and two-year-old plants lost foliage after such treatment but remained alive and developed buds four weeks after treatment. Herbicide treatment of plants with the bar gene did not cause significant deviations in height (first year) or the accumulation of aboveground biomass (second year). The obtained results improve our understanding of the effect of PPT on woody plants and can be used both to clarify mechanisms of herbicide action and in plantation forestry.