The present study investigates risk factors for onset of Clostridium difficile-associated diarrhoea, specific ribotype and environmental spore contamination in a District General Hospital in South East England. C. difficile isolates were ribotyped from 97 diarrhoeal cases, following detection of C. difficile toxin from faecal specimens by enzyme immunoassay (Health Protection Agency, Southampton). The isolates were tested for various antimicrobial susceptibilities by E-test. Cases were assessed for prior antibiotic use and followed up for clinical outcomes. Controls were matched for age, sex, ward, length of stay and comorbidity to identify any antibiotic risk factors using conditional logistic regression analysis. Environmental sampling on wards was performed with cycloserine-cefoxitin-egg yolk agar. Forty-five percent C.difficile isolates ribotyped as 027, 39% as 106 and 10% as 001. All ribotypes were resistant to ciprofloxacin, erythromycin and cefotaxime but remained susceptible to metronidazole and vancomycin. The crude (death within 28 days) and early (death within 72h) mortalities were 23% and 11% for the 027 strain, whereas for the 106 ribotype they were 11% and 3%, respectively. The case-control study identified ciprofloxacin usage for >7 days as a significant risk factor (adjusted odds ratios of 3.72; 95% CI: 1.38-10.02; P=0.019). Environmental sampling revealed the presence of spores on faecally contaminated equipment such as commodes and bedpan shells, which persisted after cleaning. Ciprofloxacin appears to encourage C.difficile-associated diarrhoea and should be restricted to short courses. Cleaning agents for clinical equipment must have sporicidal activity to prevent cross-transmission.