An important feature of tryptophan phosphorescence, crucial for probing protein structure and dynamics, is the drastic reduction of the lifetime (tau) in fluid solutions. Initial reports of indole and derivatives showed that tau decreases from 6 s in rigid glasses to about 1 ms in aqueous solutions at ambient temperature. Recently a report by Fischer et al. questioned the validity of the millisecond lifetime, claiming that in millimolar electrolyte solutions tau is about 40 micros, similar to the 12-30 micros of earlier determinations based on flash photolysis. Longer lived phosphorescence was detected in pure water but because it exhibited an initial growing phase and an anomalously large triplet yield, the emission was attributed to an artifact arising from the slow, first-order, geminate recombination of the radical cation and electron generated by photochemistry. In this study, we reexamine both the phosphorescence lifetime and the triplet quantum yield of indole, N-acetyl tryptophanamide (NATA), N-methyl tryptophan and the tryptophan-glycine-glycine tripeptide under the same conditions adopted by Fischer et al. as well as over a wider range of electrolyte and buffering salts concentrations, pH, solvent and temperature. Throughout, the results show that the phosphorescence decay is slow and uniform down to the 12 micros resolution of the instrument, with no evidence of short-lived, 40 micros-like components. Most compelling was the similarity between the fluorescence-normalized triplet yield of indole derivatives in water and that of W59 in the protein ribonuclease T1 or of NATA in rigid glasses. Its invariance over experimental conditions that varied the production of photoproducts several fold and the characteristic susceptibility of the triplet lifetime to O2, proton and ground state quenching demonstrated that the triplet state was formed predominantly through normal intersystem crossing and that its unquenched lifetime was at least 9 ms.