We describe the adaptation and evaluation of an error-compensating met
hod for kinetic determinations of deoxyribonucleic acids (DNAs). The D
NA is first reacted with ethidium bromide to produce a fluorescent int
ercalation complex. Subsequent treatment of the complex with DNase cat
alyzes hydrolysis of the DNA, causing a time-dependent decrease in flu
orescence, which is monitored. A model for two-component parallel firs
t-order processes is fit to the decay curve to predict the total chang
e in fluorescence expected if the process were monitored to equilibriu
m. The predicted change in fluorescence response varies linearly with
DNA concentration with an intercept corresponding to 0.13 mg/L DNA. Re
sults by the predictive method are 47-, 58-, and 250-fold less depende
nt on DNase activity, temperature, and ethidium bromide concentration,
respectively, than are results for an initial-rate method utilizing t
he same data. Moreover, the predictive method yields a significantly w
ider linear range than the initial-rate method, and is much less affec
ted by blank fluorescence and RNA interference than is an equilibrium
method based on the reaction of DNA with ethidium bromide alone.