Introduction Oral targeted therapies (OTTs) are widely used for cancer management. However, there is no consensus on OTT dose adaptation in older patients with cancer. Methods This noninterventional, retrospective study was a real‐life assessment of dose adaptation for six OTTs (afatinib, everolimus, palbociclib, pazopanib, sorafenib and sunitinib), at baseline and during treatment, and the reasons for the changes, in ≥70‐year‐old patients treated between February 2016 and August 2019. Data were compared with univariate models fitted with all variables. Results Among the 986 patients treated with OTT, the group of ≥70‐year‐old patients (n = 122) received afatinib (15.6%), everolimus (14.8%), palbociclib (50.8%), pazopanib (9.8%), sorafenib (5.8%) or sunitinib (3.2%). At baseline, the prescribed OTT dose was adapted (reduction) in 29% of ≥70‐year‐old patients (35/122). These 35 patients were significantly older (mean age 80 vs 74 years, P < .001), and more frequently had a performance status score ≥2 (P < .01) than the other patients (n = 87). In the standard dose group, higher toxicity grades (P = .18) and subsequent dose reduction (41% of patients, 36/87) tended to be more frequent compared with the baseline adapted dose group (26%, 9/35, P = .1). At the study end, 53% of patients in the whole cohort (65/122) were taking a lower dose than the recommended one. Conclusion At OTT initiation, dose was adapted in 29% of older adults with cancer, rarely after a formal oncogeriatric evaluation (6.5% of all patients). In the absence of recommendations, clinical studies are needed to evaluate the efficacy and safety of baseline OTT dose reduction in older adults with cancer.