Many bacterial pathogens secrete potent toxins to aid in the destruction of host tissue, to initiate signaling changes in host cells or to manipulate immune system responses during the course of infection. Though methods have been developed to successfully purify and produce many of these important virulence factors, there are still many bacterial toxins whose unique structure or extensive post-translational modifications make them difficult to purify and study in in vitro systems. Furthermore, even when pure toxin can be obtained, there are many challenges associated with studying the specific effects of a toxin under relevant physiological conditions. Most in vitro cell culture models designed to assess the effects of secreted bacterial toxins on host cells involve incubating host cells with a one-time dose of toxin. Such methods poorly approximate what host cells actually experience during an infection, where toxin is continually produced by bacterial cells and allowed to accumulate gradually during the course of infection. This protocol describes the design of a permeable membrane insert-based bacterial infection system to study the effects of Streptolysin S, a potent toxin produced by Group A Streptococcus, on human epithelial keratinocytes. This system more closely mimics the natural physiological environment during an infection than methods where pure toxin or bacterial supernatants are directly applied to host cells. Importantly, this method also eliminates the bias of host responses that are due to direct contact between the bacteria and host cells. This system has been utilized to effectively assess the effects of Streptolysin S (SLS) on host membrane integrity, cellular viability, and cellular signaling responses. This technique can be readily applied to the study of other secreted virulence factors on a variety of mammalian host cell types to investigate the specific role of a secreted bacterial factor during the course of infection.