The behavior and performance of four cloud condensation nucleus instruments are theoretically analyzed. They include the static diffusion cloud chamber (SDCC), the Fukuta continuous flow spectrometer (FCNS), the Hudson continuous flow spectrometer (HCNS), and the California Institute of Technology continuous flow spectrometer (CCNS). A numerical model of each instrument is constructed on the basis of a general fluid dynamics code coupled to an aerosol growth/activation model. Instrument performance is explored by simulating instrument response when sampling a monodisperse ammonium sulfate aerosol. The uncertainty in the wall temperature boundary condition is estimated for all the instruments and is found to be appreciable only for the CCNS. The CCNS and HCNS models reasonably reproduced experimental data, while reported limits were also verified by the FCNS model. Regarding the performance of each instrument, simulations show that the SDCC produces droplets that are monodisperse to within 10% of the particle diameter (for particles of a constant critical supersaturation). The FCNS can potentially activate particles over a wide range of critical supersaturations, but the prevailing design exhibits low sensitivity to particles with critical supersaturations below 0.1% as a result of the short time available for droplet growth under low supersaturations. The resolution capability of both HCNS and CCNS with respect to critical supersaturation is shown to be particularly sensitive to operational parameters. This is a consequence of the strongly nonlinear nature of droplet growth; droplet size cannot always be used to distinguish particles with different critical supersaturation because of the growing droplets' trend toward monodispersity. Of the two instruments, the HCNS generally displays higher resolution capability. This is attributed to the smoother and monotonic supersaturation profiles established in the HCNS. While different design parameters or operating conditions may lead to modest shifts in the performance from that predicted here for any of the four instruments, the essential features described in this paper are inherent to their designs. ¿ 2001 American Geophysical Union