In this study, four different types of mould‐ripened Civil cheese were manufactured. A defined (nontoxigenic) strain of a Penicillium roqueforti (SC 509) was used as secondary starter for the manufacture of mould‐ripened Civil cheese with and without addition of the whey cheese Lor; in parallel, secondary starter‐free counterparts were manufactured. A total of 83 compounds were identified. Ketones, alcohols and esters were the principal classes of volatile components. Principal component analysis of the headspace volatiles grouped cheeses by age and type. P. roqueforti inoculated cheese was clearly separated from the other cheeses at 180 days of ripening, and these cheeses were characterised with high levels of ketones (e.g., 2‐butanone, 2‐heptanone). Differences in the panel scores between the cheese samples were not significant during the first stage of ripening (up to 60 days); as ripening proceeded, these differences were become evident and P. roqueforti inoculated cheeses received higher scores than others. Addition of Lor in the manufacture of mould‐ripened Civil cheese caused lower points by the sensory panel, and the cheese inoculated with P. roqueforti and Lor‐free was the best type of mould‐ripened Civil cheese. The results showed that the use of P. roqueforti in the manufacture of mould‐ripened Civil cheese has significant impact on the volatile profiles and sensory attributes.