The activity of individual microorganisms can be measured within environmental samples by detecting uptake of isotope‐labelled substrates using nano‐scale secondary ion mass spectrometry (nanoSIMS). Recent studies have demonstrated that sample preparation can decrease 13C and 15N enrichment in bacterial cells, resulting in underestimates of activity. Here, we explore this effect with a variety of preparation types, microbial lineages and isotope labels to determine its consistency and therefore potential for correction. Specifically, we investigated the impact of different protocols for fixation, nucleic acid staining and catalysed reporter deposition fluorescence in situ hybridization (CARD‐FISH) on >14 500 archaeal and bacterial cells (Methanosarcina acetivorans, Sulfolobus acidocaldarius and Pseudomonas putida) enriched in 13C, 15N, 18O, 2H and/or 34S. We found these methods decrease isotope enrichments by up to 80% – much more than previously reported – and that the effect varies by taxa, growth phase, isotope label and applied protocol. We make recommendations for how to account for this effect experimentally and analytically. We also re‐evaluate published nanoSIMS datasets and revise estimated microbial turnover times in the marine subsurface and nitrogen fixation rates in pelagic unicellular cyanobacteria. When sample preparation is accounted for, cell‐specific rates increase and are more consistent with modelled and bulk rates.