Reducing falls in hospital requires an environmental as well as a patient-orientated approach. We studied patient and ward characteristics relating to falls in an acute setting. In a prospective open observational study, we examined fall characteristics in two nuclear designed wards (A and B) and a longitudinal ward (C). We recorded 63 falls among 1609 patients. Ward C had the most falls (31 vs 18/14; p = 0.01), fall positive days (29 vs 15/10; p = 0.002) and fallers (27 vs 13/12; p = 0.001; OR 2.54, CI--1.41-4.57). Ward C had a higher cumulative risk of falls (p = 0.006) and fall positive days (p = 0.003). Choice of ward was a significant independent risk factor for falls (p = 0.01) when controlled for age, sex, and diagnostic variation between the wards. Most falls were intrinsic (A 66.7%, B 64.2%, C 61.3%, p = 0.45). A significantly higher proportion of falls on ward C occurred by the bed (p = 0.04). Significant differences exist between the wards, and fall reduction programmes should identify and compensate for adverse ward-related factors to increase the effectiveness of patient-targeted fall risk assessments.