The R programming language is now widely used, and taught, for data processing in psychology and many social sciences. Despite its popularity and the widespread familiarity with R among students and researchers, when it comes to building experiments, both groups have had to pick up entirely new programming languages and skillsets. The glia package aims to remedy this, providing an accessible toolkit for constructing and running experiments from R. However, R was not originally designed to meet the demands of low-latency interaction and fast stimulus presentation that are fundamental to laboratory research. Therefore, we investigate whether, and how, it can be adapted, to combine a familiar and accessible programming language with the needs of scientists in terms of presentation latency and response timing accuracy. We demonstrate that the glia package can achieve high-fidelity timing from R, validate its performance with regard to stimulus presentation and response time measurement, and present a general pattern for building performant experiments in environments not built for this purpose.