1.Ecosystem processes and the organisms involved are generally limited by the availability of one or more element in soil, an important phenomenon to consider for our understanding of ecosystem functioning and future changes. Especially in tropical forests, typically growing on nutrient depleted soils, nitrogen (N), phosphorus (P) or other limitations are assumed. However, large‐scale nutrient manipulation experiments revealed complex site‐specific patterns and several authors raised the need for novel approaches to reveal deeper mechanistic insights on limitation patterns, especially concerning soil microbial activity. 2.In order to gain such deeper knowledge, based on a short review of previous small‐scale studies focusing on soil microorganisms, we developed an experimental approach which controls for common biases, including indirect treatment effects, addition of co‐elements or nutrient pulses. Using this technique, we tested the hypotheses that fungi growing in tropical forest soils are mainly limited by P, and that N versus P limitations shift along an altitudinal gradient. 3.Mesh bags of 38 μm filled with sand were amended with weak ion exchange resins loaded with N, P or potassium (K) and buried underneath the litter layer at three altitudinal sites in Southern Ecuador. After a period of four months, the in‐growth of fungal hyphae was quantified, phospholipid fatty acids were analyzed for a subset of samples and chemical properties were determined. 4.In line with the first hypothesis, hyphal abundance was increased in P amended mesh bags, indicating P limitation. However, this pattern was not significantly shifted along the altitudinal gradient. By contrast, N addition increased hyphal abundance at the lowest site, compared to significant reductions at 2000 and 3000 m – not only in fungi but also in bacterial abundance as indicated by PLFA analyses, contradicting common soil‐age hypotheses. Decreased nutrient immobilization and fungal in‐growth at higher elevations suggest slow microbial activity, including nitrification, which may have caused toxic ammonium accumulation. 5.The experimental design offers a promising tool to provide more mechanistic and soil focused analyses specifically targeting microbes, which in this system strongly supported the hypothesis of primary fungal P limitation in tropical forest soils.