Bioturbation refers to the mixing of sediment particles resulting from benthic faunal activity. It is the dominant particle mixing process in most marine sediments and exerts an important control on diagenetic processes. In models, bioturbation is usually treated as a diffusive process where the biodiffusion coefficient (Db) characterizes the biological mixing intensity. Biodiffusion coefficients are classically computed by fitting a diffusive model to vertical profiles of particle-bound radioisotopes. One peculiar observation is tracer-dependence: Db values from short-lived tracers tend to be larger than those obtained from long-lived tracers from the same site. Recent theoretical work, based on random walk theory and Lattice Automaton Bioturbation Simulations (LABS), has suggested that this tracer-dependence is simply a model artifact and has concluded that the biodiffusion model is not applicable to the short observational time scales associated with short-lived radioisotopes. Here we have compiled a global dataset of Db values obtained from different radiotracers to assess tracer-dependence from a data perspective. Tracer-dependence is significant in low-mixing environments like slope and deep-sea sediments, but is not present in intensely mixed coastal areas. Tracer-dependence is absent when the number of mixing events is larger than 20, or the potential length scale is greater than 0.5 cm. Roughly this comes down to tracer-derived Db values greater than 2 cm2 yr−1. This condition is met for 68%, 50%, and 8% of published Db values obtained from coastal, continental slope, and abyssal environments, respectively. These results show that short-lived radioisotopes are suitable to quantify biodiffusion mixing in sedimentary environments featuring intense bioturbation.