The method of decreasing conversion, constant space-time (DCCST) is sometimes used for empirical determination of catalyst decay kinetics in fixed beds. The method of constant conversion, increasing space-time (CCIST) is used infrequently. Although it is well known that these methods are most readily applied to cases in which the deactivation is independent of the composition of the fluid phase, the ease with which they can be misapplied to concentration-dependent cases has not been studied systematically. In this work, the empirical determination of catalyst decay kinetics in fixed beds has been examined by numerical simulation. It has been shown that it is rather easy to generate an invalid concentration-independent decay model, either by fitting a dubious straight line to a limited range of data using the DCCST method or by fitting a good, or even perfect, straight line using the CCIST method. It has also been shown that both methods can, in principle, be used reliably for empirical determination of catalyst decay kinetics, whatever the dependence on composition, provided that multiple experiments are carried out over a range of conversion. The CCIST method has much greater potential for resolving reaction and decay kinetics than has been realized previously, and it can be applied to reversible as well as irreversible reactions. This method will generally give a linear plot, provided that the selected plot corresponds to the actual decay order. The plot is exactly linear for any first-order decay kinetics and almost linear for non-first-order decay kinetics. Exact analytical solutions have been derived for various first-order cases. These solutions can be used to determine the true decay kinetics from multiple parallel experiments. Deviation from ideal behavior in this method is a useful indication that the power-law decay kinetics has a limited range of validity, beyond which a more detailed, systematic investigation may be required.