A number of resources are available for evaluating the rigor of single-case designs, including the commonly used multiple baseline design. In this article, we discuss two characteristics commonly cited as necessary for the highest rigor in multiple baseline designs--concurrence and response-guided baseline condition duration. We suggest that both nonconcurrence and response-independent baseline condition duration may be appropriate in some situations. Suggestions for researchers include improved reporting and graphing. Future research is needed to determine experimental differences when these design variations are used.