Organic soils have low sorption capacities for phosphorus (P), and may pose a risk of P loss to water if P applications to these soils coincide with runoff events. Little is known about the magnitude of exports of P in overland flow following application of P fertiliser onto these soils, or on the influence of the frequency on P losses and persistence. The number of P fertiliser applications was surveyed across 39 commercial farms to assess current practice and inform the design of a rainfall runoff experiment to evaluate the effect of frequency of P applications on losses and persistence across time. Superphosphate (16 % P) was applied in single (equivalent to 30 and 55 kg P ha-1 applied at day 0) and split (equivalent to 15 and 27.5 kg P ha-1 applied in two doses at days 0 and 55) applications to an organic soil inclined at a slope of 6 % in a rainfall simulator experiment. The surface runoff of dissolved reactive phosphorus (DRP) was measured in controlled 30-min rainfall simulations conducted intermittently over an 85-day period. The DRP loads in surface runoff after the first rainfall event were 44.6 and 97.8 mg L-1 for the single applications of 30 and 55 kg ha-1, respectively, and 13.3 and 21.8 mg L-1 for the same rates split in two doses, indicating that single P applications had disproportionately bigger impacts on losses than split applications. This supports the idea that frequent, but smaller, P applications can minimise the impact of fertilisation on waters. Dissolved reactive P concentrations remained significantly higher than those from the control samples until the end the experiment for almost all the P treatments, highlighting the long-lasting effects of added P and the elevated risk of P losses on organic soils. For climates with frequent rainfall events, which are likely to coincide with fertiliser applications, smaller but more frequent P applications can reduce the risk of P transfer as opposed to one single application.