Abstract: Background: beta-delayed multiple neutron emission has been observed for some nuclei with A <= 100 being the Rb-100 the heaviest beta 2n emitter measured to date. So far only 25 P-2n values have been determined for the approximate to 300 nuclei that may decay in this way. Accordingly it is of interest to measure P-2n values for the other possible multiple neutron emitters throughout the chart of the nuclides. It is of particular interest to make such a measurement for nuclei with A > 100 to test the predictions of theoretical models and simulation tools for the decays of heavy nuclei in the region of very neutron-rich nuclei. In addition the decay properties of these nuclei are fundamental for the understanding of astrophysical nucleosynthesis processes such as the r-process and safety inputs for nuclear reactors. Purpose: To determine for the first time the two-neutron branching ratio the P-2n value for Sb-136 through a direct neutron measurement and to provide precise P-1n values for Sb-136 and Te-136. Method: A pure beam of each isotope of interest was provided by the JYFLTRAP Penning trap at the Ion Guide Isotope Separator On-Line (IGISOL) facility of the University of Jyvaskyla Finland. The purified ions were implanted into a moving tape at the end of the beam line. The detection setup consisted of a plastic scintillator placed right behind the implantation point after the tape to register the beta decays and the BELEN detector based on neutron counters embedded in a polyethylene matrix. The analysis was based on the study of the beta- and neutron-growth-and-decay curves and the beta-one-neutron and beta-two-neutron time correlations which allowed us the determination of the neutron branching ratios. Results: The P-2n value of Sb-136 was found to be 0.14(3)% and the measured P-1n values for Sb-136 and Te-136 were found to be 32.2(15)% and 1.47(6)% respectively. Conclusions: The measured P-2n value is a factor 44 smaller than predicted by the finite-range droplet model plus the quasiparticle random-phase approximation (FRDM+QRPA) model used for r-process calculations.