The azide ion is a strong physical quencher of singlet molecular oxygen (1O2) and is frequently employed to show involvement of 1O2 in oxidation processes. Rate constants (k(q)) for the quenching of 1O2 by azide are routinely used as standards to calculate k(q) values for quenching by other substrates. We have measured k(q) for azide in solvent mixtures containing deuterium oxide (D2O), acetonitrile (MeCN), 1,4-dioxane, ethanol (EtOH), propylene carbonate (PC), or ethylene carbonate (EC), mixtures commonly used for many experimental studies. The rate constants were calculated directly from 1O2 phosphorescence lifetimes observed after laser pulse excitation of rose bengal (RB), used to generate 1O2. In aqueous mixtures with MeCN and carbonates, the rate constant increased nonlinearly with increasing volume of organic solvent in the mixtures. k(q) was 4.78 x 10(8) M(-1) s(-1) in D2O and increased to 26.7 x 10(8) and 27.7 x 10(8) M(-1) s(-1) in 96% MeCN and 97.7% EC/PC, respectively. However, in EtOH/D2O mixtures, k(q) decreased with increasing alcohol concentration. This shows that a higher solvent polarity increases the quenching efficiency, which is unexpectedly decreased by the proticity of aqueous and alcohol solvent mixtures. The rate constant values increased with increasing temperature, yielding a quenching activation energy of 11.3 kJ mol(-1) in D2O. Our results show that rate constants in most solvent mixtures cannot be derived reliably from k(q) values measured in pure solvents by using a simple additivity rule. We have measured the rate constants with high accuracy, and they may serve as a reliable reference to calculate unknown k(q) values.