Background: Low-carbohydrate diets are becoming increasingly popular, although their dietary quality outside of clinical studies is unknown. A previous study analysed the dietary intake in people consuming a reduced-carbohydrate diet (<40% calories). However, it is not clear what foods people consume when carbohydrate is reduced to below 26% of total calories. Methods: In the present cross-sectional study, the dietary and nutrient intake collected via up to five consecutive 24-h dietary recalls and a food frequency questionnaire of 444 individuals (aged 46-79 years) consuming <26% of calories from carbohydrate (LCHO) was compared with that of 131 897 individuals consuming ≥45% calories from carbohydrate (NCHO) using the UK Biobank Dataset. Absolute cut-offs to define the low-carbohydrate group (<130 g day–1; n = 1953 versus ≥225 g day–1, n = 113 036) were also used. Results: Both NCHO (>45% calories and ≥225 g) groups consumed significantly more high-sugar, high-fat snacks [median 6.0, interquartile range (IQR) = 2.0–11.0 and median 6.0, IQR = 3.0–11.8, respectively) compared to the LCHO (<26% calories and <130 g) groups (median 0, IQR = 0–2.8 and median 1, IQR = 0–3.8, respectively) (P < 0.0001). Both LCHO groups reported consuming significantly more red meat, oily fish, nuts and seeds but fewer fruits, vegetables and pulses compared to the NCHO groups. In general, the consumption of oily fish, nuts, seeds and pulses was low across the whole cohort and differences in intake between the LCHO and NCHO groups were small. After adjusting for socio-economic status, most differences remained. Conclusions: Carbohydrate restriction is associated with both beneficial and potentially deleterious dietary changes compared to a normal carbohydrate intake.