One measurement and an algebraic formula are used to calculate the incident air kerma (Ka,i) at the skin after any CT examination, including cone-beam CT (CBCT) and multi-slice CT (MSCT). Empty scans were performed with X-ray CBCT systems (dental, C-arm and linac guidance scanners) as well as two MSCT scanners. The accumulated Ka,i at the flat panel (in CBCT) or the maximum incident air kerma at the isocentre (in MSCT) were measured using a solid-state probe. The average Ka,i(skin), at the skin of a hypothetical patient, was calculated using the proposed formula. Additional measurements of dose at the isocentre (DFOV) and kerma-area product (KAP), as well as Ka,i(skin) from thermoluminiscence dosimeters (TLDs) and size-specific dose estimates are presented for comparison. The Ka,i(skin) for the standard head size in the dental scanner, the C-arm (high dose head protocol) and the linac (head protocol) were respectively 3.33 ± 0.19 mGy, 15.15 ± 0.76 mGy and 3.23 ± 0.16 mGy. For the first MSCT, the calculated Ka,i(skin) was 13.1 ± 0.7 mGy and the TLDs provided a Ka,i(skin) between 10.3 ± 1.1 mGy and 13.8 ± 1.4 mGy. Estimation of patient air kerma in tomography with an uncertainty below 7% is thus feasible using an empty scan and conventional measurement tools. The provided equations and website can be applied to a standard size for the sake of quality control or to several sizes for the definition of diagnostic reference levels (DRLs). The obtained incident air kerma can be directly compared to the Ka,i from other X-ray modalities as recommended by ICRU and IAEA.