Standard procedures to assess P availability in soils are based on batch experiments with various extractants. However, in most soils P nutrition is less limited by bulk stocks but by strong adsorption and transport limitation. The basic principle of root‐phosphate uptake is to strip phosphate locally from the solid phase by forming a radial depletion zone in the soil solution, optionally enhanced by release of mobilizing substances. Microdialysis (MD), a well‐established method in pharmacokinetics, is capable to mimic important characteristics of P root uptake. The sampling is by diffusional exchange through a semipermeable membrane covering the probes with their sub‐mm tubular structure. Additionally, the direct environment of the probe can be chemically modified by adding, e.g., carboxylates to the perfusate. This study is the first approach to test the applicability of MD in assessing plant available phosphate in soils and to develop a framework for its appropriate use.We used MD in stirred solutions to quantify the effect of pumping rate, concomitant ions, and pH value on phosphate recovery. Furthermore, we measured phosphate yield of top‐soil material from a beech forest, a non‐fertilized grassland, and from a fertilized corn field. Three perfusates have been used based on a 1 mM KNO3 solution: pure (1), with 0.1 mM citric acid (2), and with 1 mM citric acid (3). Additionally, a radial diffusion model has been parametrized for the stirred solutions and the beech forest soil.Results from the tests in stirred solutions were in good agreement with reported observations obtained for other ionic species. This shows the principal suitability of the experimental setup for phosphate tests. We observed a significant dependency of phosphate uptake into the MD probes on dialysate pumping rate and on ionic strength of the outside solution. In the soils, we observed uptake rates of the probes between 1.5 × 10−15 and 6.7 × 10−14 mol s−1 cm−1 in case of no citrate addition. Surprisingly, median uptake rates were mostly independent of the bulk soil stocks, but the P‐fertilized soil revealed a strong tailing towards higher values. This indicates the occurrence of hot P spots in soils. Citrate addition increased P yields only in the higher concentration but not in the forest soil. The order of magnitude of MD uptake rates from the soil samples matched root‐length related uptake rates from other studies. The micro‐radial citrate release in MD reflects the processes controlling phosphate mobilization in the rhizosphere better than measurements based on “flooding” of soil samples with citric acid in batch experiments. Important challenges in MD with phosphate are small volumes of dialysate with extremely low concentrations and a high variability of results due to soil heterogeneity and between‐probe variability. We conclude that MD is a promising tool to complement existing P‐analytical procedures, especially when spatial aspects or the release of mobilizing substances are in focus.