Trace elements are often omitted from chemically defined growth media. We show that this omission makes experiments unnecessarily sensitive to unavoidable contamination with trace elements. To investigate this experimentally, we grow eleven bacterial strains in high replicate with and without supplementing trace elements, keeping all other conditions as fixed as possible. Even in this benign scenario, we find that supplementing trace elements considerably reduces variability of growth. We discuss implications for the design and use of trace-element supplements, and in particular argue that using them increases experimental reproducibility and should be standard practice.