The absolute second-order reaction rate coefficient, k(1), for the gas phase reaction, O(P-3) + N2H4 --> products, was studied in a discharge flow-tube apparatus. The reaction was studied under pseudo-first-order conditions in O(P-3) concentration (i.e., [N2H4] much greater than [O(P-3)]). The O atoms were generated by a microwave discharge of a suitable precursor gas in He in a fixed side-arm reactor upstream of the flow tube, or in the sliding inner injector of the flow tube. The hydrazine concentration was photometrically measured and introduced into the apparatus in a flow of He via the sliding injector or the fixed side-arm port, respectively. The kinetics of the O-atoms in the reaction was directly followed by 130.2-130.6 nm cw-resonance fluorescence detection of O(P-3) at the fixed detector situated downstream of the flow tube. The Arrhenius expression, k(1) = (7.35 +/- 2.16) x 10(-13) exp[(640 +/- 60)/T] cm(3) molec(-1) s(-1), in the temperature range 252 - 423 K, was fit to the data points. The rate coefficient at room temperature was, within experimental errors, independent of the He buffer gas pressure in the range 1.74 to 8.30 Torr, or the O-atom source reactor. The formation of OH(X(2) Pi) in the reaction, which can be vibrationally excited (v" less than or equal to 2), was directly detected by pulsed laser-induced fluorescence. The total yield of OH in the reaction was determined to be (0.15 +/- 0.05) at 298 K, of which similar to 50% is thought to be produced vibrationally hot. These results suggest that the single-H-atom removal channel is a minor process, in agreement with earlier molecular beam studies in which a direct two-H-atom removal channel was proposed to be the principal reaction mechanism by which O(3P) reacts with N2H4.