In priciple dont care to much- both are a kind of relativ unit. cps on the fluoromax can give you a hint, if you detector is going to saturate because of high intensity. One question- are you measuring only emssion specta on the same excitation wavelength? Otherwise it could be a good idea to normalise to the excitation channel- that would give you also AU and a better comperability.

Although counts per second sounds like an absolute measurement, it is still arbitrary because it is determined by instrument settings, such as wavelength, band width and gain. A non-arbitrary measurement would be quantum yield.

Jan Voigt i am measuring only emssion spectra on the same excitation wavelength

I am not sure- if you got cps on a fluoromax, did you get the (by sensitivity of detection channel) corrected spectra? It is been some days ago, that I used a fluorolog.

 If a photon-counting instrument is used to obtain the spectrum, one often notes that the Y-axis is labeled counts per second or cps.  Sometimes, if an analog instrument is utilized, the Y-axis is labeled as “voltage,” although this convention is rare.  Quite often, the Y-axis is given in “arbitrary units.”  Why are emission (and excitation) intensities reported in this manner?  Absorption values have specific units, namely absorbance or optical density, which are independent of instrument parameters such as lamp intensity, slitwidths, PMT voltages and amplifier gains, since the measurements are always referred to the intensity of the incident light.  In the case of fluorescence, however, the observed fluorescence intensity will depend upon how much light illuminates the sample – which is a function of the lamp power, the excitation slitwidth, and the efficiency of the excitation monochromator.  The observed intensity will also depend on the emission slitwidth, the efficiency of the emission monochromator, the PMT voltage, the details of the electronic amplification (e.g. analog versus photon counting) and the particular optical arrangement of the instrument, i.e., the focal length of the lenses and even the nature of the cuvette utilized, i.e., pathlength.  These parameters vary from instrument to instrument, and sometimes even from measurement to measurement.  Thus, even though the signal being recorded may actually be the number of photons detected, these counts will vary with extrinsic instrumentation parameters in addition to the fluorophore concentration and photophysical parameters (e.g., extinction coefficient and quantum yield) and hence the descriptor “arbitrary” is utilized.  It is also common to see the fluorescence intensity normalized to unity, i.e., the maximum is set to “1.”