A short collaborative study that is designed to begin an experimental assessment of variability in radiocarbon dating is described. An identical set of eight wood samples, taken from across disks of the same tree, was distributed to, and dated independently by, each of 20 ¹⁴C laboratories. The results show order-of-magnitude agreement in ¹⁴C age but also reveal the existence of systematic bias and unexplained variability in the data from many labs.