Fluorescent intercalating and groove binding dyes that associate with double stranded DNA (dsDNA) are used for detection of hybridization in some sensor and biochip designs. It is possible that dye–dye interactions at concentrations of dye that are relevant to biosensor use can lead to unexpected and undesired emission wavelength shifts and fluorescence quenching interactions. A specific example of such a situation is described in this work, and suggests cautions and considerations for development of quantitative biosensors and biochips. It has been reported [I.I. Timcheva, V.A. Maximova, T.G. Deligeorgiev, N.I. Gadjev, R.W. Sbnis, I.G. Ivanov, FEBS Lett. 405 (1997) 141] that 1-hydroxyethyl-4-[(3-methyl-6-methoxybenzothiazole-2-ylidine)methine] quinolium bromide (TOMEHE), a derivative of thiazole orange, may be useful for quantitative determination of DNA hybridization. In this manuscript, we report evaluation of TOMEHE for use as a fluorescent indicator of DNA hybrid formation on a fiber optic genosensor platform. The wavelength of maximum fluorescence emission of TOMEHE associated with immobilized single-stranded polythymidylic acid isocanucleotides (dT 20) on the surface of the optical fiber was determined to be 560±3 nm and was not observed to be significantly different from that in the presence of dsDNA, contrary to the initial report (Timcheva et al., 1997). At high dye concentrations, a bathochromic shift in the fluorescence emission intensity maximum of TOMEHE was observed with both single stranded DNA (ssDNA) and dsDNA, and was believed to be the result of self-association of fluorophore. TOMEHE demonstrated a 10-fold greater fluorescence intensity over ethidium bromide (EB) when bound to dsDNA, but also displayed a 2.5-fold increase in fluorescence intensity over EB when associated with ssDNA. To maximize S/ N, many biosensor and biochip experiments use dye concentrations that are in large excess in comparison to the quantity of immobilized DNA. The linearity of fluorescence intensity response for TOMEHE when bound to dsDNA was poor ( r 2=0.566) over the range studied (1:60 to 1:1 dye:base pair ratios), but improved at dye:base pair ratio ranges below 1:12.