In this 1 year cross-sectional study of personnel being discharged from compulsory military service, an available database of health-related information was used to examine the association of meningococcal carriage with socio-demographic factors. A representative, systematic sample of 1632 personnel was interviewed and had throat cultures taken. The overall meningococcal carriage rate was 16%. Serogroups B and Y accounted for 76% and 13% of the isolates respectively. In univariate analysis, carriage was associated with male gender (P < 0.0001), < 12 years school education (P = 0.002), smoking (P = 0.014), and service at a 'closed' base, reflecting greater interpersonal contact (P < 0.0001). In multivariate analysis, only service on a closed base and male gender retained significance. School education of < 12 years remained significant for females only. Variables not associated with carriage included number of siblings, intensity of smoking, and use of the contraceptive pill. In this setting, meningococcal carriage was associated with the type of base on which soldiers served; and smoking was not an independent risk factor for carriage.