Objective: Assessment of consumption of raw and processed tomatoes, plasma lycopene concentration and their interrelationship in a nationally representative sample of elderly British people. Setting: A 1994-1995 National Diet and Nutrition Survey: a nationwide cross-sectional sample of people aged 65 y and over living in mainland Britain, either in the community ('free-living', n ¼ 1275) or in institutions such as nursing homes (n ¼ 412). Methods: Consumption of raw and processed tomatoes (including those in tomato-containing composite foods) was recorded by a 4-day dietary record for each participant. Plasma lycopene concentration was assayed by HPLC. Sociodemographic and lifestyle information was recorded by trained interviewers. Results: In all, 29% of free-living people and 24% of those living in institutions did not consume any raw or processed tomatoes during the 4-day recording period. Free-living participants were more likely to consume raw tomatoes (26 vs 16%). Plasma lycopene concentration was positively associated with the weight of raw and processed tomatoes consumed, especially with processed tomatoes and their products, alcohol consumption and with plasma total (as well as HDL and non-HDL) cholesterol concentrations. Among free-living participants, increased age, cigarette smoking habit and manual occupational social class were all associated with lower plasma lycopene concentration. Conclusions: In the current population, consumption of raw and processed tomatoes varied by sociodemographic and lifestyle characteristics, and made a significant contribution to plasma lycopene concentration. Determinants of plasma lycopene concentration were age, plasma cholesterol concentration and smoking habit.