Background Although tissue eosinophilia has traditionally been considered diagnostically supportive of adverse cutaneous drug reactions (ACDRs), studies have suggested it is neither a sensitive nor a specific finding in drug eruptions (DEs). Objectives Determining whether skin tissue eosinophilia is a reliable indicator of ACDR. Methods A nested case–control retrospective study conducted in a cohort of 170 patients at a single institution. Tissue eosinophilia (number of eosinophils per high‐power field (HPF)) was investigated in skin biopsies obtained from the following groups of patients who demonstrated: (i) in vitro assay and telephone interview‐validated cutaneous drug reactions (true DE); (ii) initial clinical diagnosis of ACDR but drug aetiology was excluded by in vitro assay and telephone interview (false DE); and (iii) non‐drug‐associated cutaneous eruptions, skin tumours and nevi, randomly selected for evaluation (control). Results Significantly higher number of eosinophils per HPF was observed in the false DE compared to the true DE group (P = 0.02). The false DE group demonstrated a higher number of eosinophils (P < 0.001) while the true DE group eosinophils’ number was not significantly higher as compared to control (P = 0.2032). Conclusions Tissue eosinophilia is not a reliable indicator of ACDRs.