Reducing soil nutrient availability has been proposed as a strategy to favor native vs. non‐native invasive plant species and represents a potential alternative to traditional manual removal or chemical control methods. We implemented a field experiment in invaded dry and wet montane Hawaiian ecosystems to test responses of soil and dominant plant species to three soil nutrient treatments (Control = no nutrient manipulation; Carbon = C substrate added to reduce nutrients; Fertilizer = fertilizer added to increase nutrients) and two non‐native plant treatments (Weeds removed; Weeds present) in a fully factorial experiment in each ecosystem over 18 months. Carbon amendments reduced soil inorganic nutrient availability by 60–70% in dry shrubland and 30–50% in wet forest. Fertilizer amendments increased soil inorganic nutrient availability by >20‐fold. Altered nutrient availability did not impact gross mineralization or nitrification rates in either ecosystem. In dry shrubland, neither C amendments nor weed removal altered growth or reproduction, but fertilizer increased woody growth and forb/grass reproduction in both natives and non‐natives. In wet forest, weed removal but not C amendments increased growth and survival of native woody seedlings, while fertilizer decreased native seedling survival and increased non‐native woody seedling growth. Overall, growth and reproduction of native and non‐natives responded similarly to altered nutrient availability, indicating that for the tropical ecosystems and species examined, manipulating nutrient availability does not favor native versus non‐native invasive plants in the first 18 months. In contrast, weed removal had positive effects on native plant growth, likely mediated through changes in other resources.