Electronic passivation of III-V surfaces is essential for applications in optoelectronic devices. A key aspect is the measurement of the surface recombination properties which can be done by various techniques including transient photoluminescence (TRPL) or luminescence quantum yield. These measurement techniques are always indirect since they suppose postulating a mathematical model for the data interpretation. Most common models use the notion of surface recombination velocity to quantify the surface recombination. In this paper we demonstrate on InP substrate than this notion is not always sufficient to represent the injection dependence of the surface phenomena. The study of power dependence of TRPL decays coupled with Modulated Photoluminescence (MPL) spectra on four samples from the same wafer with different surface treatments (epi-ready, freshly cleaned, after air exposure and with poly-phosphazen passivation) allows us for discriminating between bulk and surface properties. We introduce surface defect trapping as an alternative to explain TRPL decays and MPL phase excitation power dependences of the three non-passivated samples. Surface trap parameters such as capture cross section and defect density are extracted. The passivated sample exhibits invariant response shape on eight orders of magnitude of illumination. The stability of the PL response at high flux is in agreement with the perfect stability of the passivation layer which is able to protect the InP surface without chemical changes over more than one year. They are linked to the nature of the Passivation layer/InP interface. Other surface treatments were found to have an injection dependent responseThe following article has been accepted by the Journal of Applied Physics.