RNA-based devices controlling gene expression bear great promise for synthetic biology, as they offer many advantages such as short response times and light metabolic burden compared to protein-circuits. However, little work has been done regarding their integration to multilevel regulated circuits. In this work, we combined a variety of small transcriptional activator RNAs (STARs) and toehold switches to build highly effective AND-gates. To characterize the components and their dynamic range, we used an Escherichia coli (E. coli) cell-free transcription-translation (TX-TL) system dispensed via nanoliter droplets. We analyzed a prototype gate in vitro as well as in silico, employing parametrized ordinary differential equations (ODEs), for which parameters were inferred via parallel tempering, a Markov chain Monte Carlo (MCMC) method. On the basis of this analysis, we created nine additional AND-gates and tested them in vitro. The functionality of the gates was found to be highly dependent on the concentration of the activating RNA for either the STAR or the toehold switch. All gates were successfully implemented in vivo, offering a dynamic range comparable to the level of protein circuits. This study shows the potential of a rapid prototyping approach for RNA circuit design, using cell-free systems in combination with a model prediction.