Background:Our previous survey on first-in-human trials (FIHT) of monoclonal antibodies (mAbs) showed that, due to their limited toxicity, the recommended phase II dose (RP2D) was only tentatively defined.Methods:We identified, by MEDLINE search, articles on single-agent trials of mAbs with an FIHT included in our previous survey. For each mAb, we examined tested dose(s) and dose selection rationale in non-FIHTs (NFIHTs). We also assessed the correlation between doses tested in the registration trials (RTs) of all FDA-approved mAbs and the corresponding FIHT results.Results:In the 37 dose-escalation NFIHTs, the RP2D indication was still poorly defined. In phase II-III NFIHTs (n=103 on 37 mAbs), the FIHT RP2D was the only dose tested for five mAbs. For 16 mAbs, only doses different from the FIHT RP2D or the maximum administered dose (MAD) were tested and the dose selection rationale infrequently indicated. In the 60 RTs on 27 FDA-approved mAbs with available FIHT, the FIHT RP2D was tested only for two mAbs, and RT doses were much lower than the FIHT MAD.Conclusions:The rationale beyond dose selection in phase II and III trials of mAbs is often unclear in published articles and not based on FIHT data. © The Author(s) named above 2018.