223Ra is a radiopharmaceutical used as unsealed source in nuclear medicine. In the case of staff inhalation contamination of 223Ra, methods to estimate the committed effective dose should be chosen with care. Three methods are available: whole-body measurement and gamma spectrometry for urine or faeces samples. Considering the analytical performances and uncertainties of these three methods, we propose recommendations for special dose assessment. As a first choice, due to its rapidity and its non-invasiveness, an in vivo analysis (with HPGe detector) is the most appropriate method. However, after 24 h, whole-body counting is not sensitive enough to detect a minimum effective dose of 1 mSv. Sufficient sensitivity can only be reached up to 8 days after contamination by true 24 h faeces samples analyses. Thus, despite its main drawbacks, this method appears to be more appropriate than urine to estimate the committed effective dose in addition to whole-body counting.