Superheated nanodroplets (NDs) were recently proposed for <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in vivo</i> proton range verification, owing to their ability to vaporize into echogenic microbubbles (MBs) upon exposure to ionizing radiation. In a previous publication, vaporization events were detected with 2D Ultrasound Localization Microscopy (ULM) based on a pulse-echo method. Here, we introduce P-ULM, a passive version of ULM, based on the detection of acoustic signatures emitted by vaporizing NDs, without actively transmitting ultrasound. Due to the lack of a time reference for the trigger of the ND vaporization, the time differences of arrival to each transducer element are used to retrieve the position of vaporizing NDs. P-ULM, compared to ULM, can continuously detect and super-localize sparse radiation-induced vaporization events with inherent specificity against already existing microbubbles, which otherwise would hinder range verification in the presence of vascular flow. We evaluated the localization performances of both methods theoretically and experimentally, by interleaving active and passive acquisitions on ND-phantoms irradiated with protons. P-ULM offered a higher sensitivity to vaporizations, as it detected twice as many events as ULM. Both methods retrieved, in the acoustic lateral direction, the proton range and range dispersion with sub-millimeter accuracy. In the acoustic axial direction, despite a degraded theoretical resolution limit, P-ULM retrieved the proton spot size with an accuracy similar to ULM. Importantly, P-ULM detected vaporization events with high specificity in the presence of flowing MBs, which makes the technique a candidate for in vivo proton range verification in the presence of flow.