Highly soluble commercial phosphorus (P) fertilisers have been the most common form of P applied in agriculture for decades, but their releasing efficiency can be low because phosphate ions can easily bind with cations in the soil to form precipitates or highly water-soluble species, leaching into nearby water sources and thus causing eutrophication. Using a coiled flow inverter (CFI), a new formulation of phosphorus fertiliser was prepared containing three main components: a solid P source (biphasic calcium phosphate), a soil remediation agent (citrate ions) and a binding agent (chitosan). Our study is the first to assess the fertiliser performance of a material made in flow that is a finished product exhibiting structural and functional complexity, and we have also demonstrated its performance by soil experiments. We provide a proof of concept at scale of this complex product, which may indicate its potential for commercial application. The key factors used to evaluate the prepared fertiliser were the P-releasing efficiency, soil-available P and P uptake. In the incubation experiment, the available P in soil applied with the prepared formulation was three times higher than soil applied with commercial apatite (probability value, p-value < 0.05) and was as high as soil applied with commercial P fertiliser. After incubating for 14 days, 63% of the applied P in the prepared fertiliser was released into the soil. The formulation without chitosan exhibited a higher uptake of P in Scepter wheat in comparison with phosphate rock. In the soil-column experiment, the commercial P fertiliser leaked double the amount of phosphorus after 23 days and had a fluctuated releasing behaviour when compared with the prepared formulation. Its batch synthesis and flow synthesis (8 mL min−1) shared similar green chemistry metrics, including an E-factor of 0.4 kg kg−1. Continuous-flow production for CS–ACP–Cit showed a higher productivity than the batch production by a factor of 1.9. The circularity of the process, denoted by the Materials Circularity Indicator (MCI), while considering experimental recycling data, was high for all the process steps and composites, scoring in the range of 0.8, which was interesting, considering that this was a non-optimised process. The circularity assessment took account of the performance of the composites, which is expressed as the intensity factor (U; performance) and lifetime (L) of the MCI.