Camelina sativa oil (CO) is characterized by a high content (up to 40 wt %) of essential α-linolenic acid and characteristic odour and flavour. Deodorization of highly unsaturated oils requires great attention as the refining process involves thermal treatment which affects oil integrity. In the present study RSM and principal component analysis (PCA) were used to optimize bench-scale deodorization of CO. Mathematical models were generated through multiple regressions with backward elimination, describing the effects of process parameters (temperature, steam flow, time) on oil quality indicators [peroxide value (PV), p-anisidine value (p-AV), γ-tocopherol (γ-T) and oxidative stability (OS)]. Additionally, sensory evaluation was performed. RSM analysis showed a significant effect of deodorization temperature and to a lesser extent, deodorization steam flow and time on removal of oxidative compounds, flavour and odour. PCA of chemical and sensory results showed that deodorization temperature affected the sensory properties in the samples. The best conditions for removing undesirable flavour and odour were achieved by using a deodorization temperature of 195–210°C.