The current paradigms for assessing fluorescence sensitivity on flow c
ytometers do not provide an adequate assessment of an instrument's abi
lity to detect and measure weak fluorescence on stained particles. The
capability to resolve dimly stained populations depends on two factor
s: the background noise (B), and the efficiency (Q) with which the flu
orescence from the fluorochrome molecules are concerted to photoelectr
ons. Any single statistical measure of fluorescence histogram distribu
tions will be unable to uniquely characterize an instrument. Therefore
, neither of the routinely used methods (detection threshold and delta
channel) measure sensitivity completely and unambiguously We show the
Limitations of these methods and propose that instrument sensitivity
be characterized in terms of both background noise and detection effic
iency in order to determine better the capability to detect and resolv
e weakly fluorescent particles. (C) 1998 Wiley-Liss, Inc.