Objective: To evaluate the measurement properties of an expanded version of the Falls Efficacy Scale (FES)-a measure of fear of falling. The original FES measures fear on almost exclusively indoor activities, which may limit the usefulness of the scale in identifying early stages of fear of falling in active community-dwelling older people. Design: Two group convenience sample. Setting: An outpatient referral clinic, and community-dwelling older people recruited from various sources. Subjects: A volunteer sample of 111 healthy community-dwelling elderly (mean age 74.0 years) and a sample of 68 older people referred to a Falls and Balance Clinic (FBC) (mean age 79.2). Twenty- one subjects (9 healthy elderly and 12 FBC patients) were tested twice 1 week apart to investigate retest reliability. Main Outcome Measures: A 14-activity questionnaire (the Modified Falls Efficacy Scale [MFESI) was used that incorporated the original 10-activity FES and four additional activities. Falls efficacy was rated on a 10-point visual analogue scale for each activity. Results and Conclusions: The MFES demonstrated high internal consistency (Cronbach's alpha .95) and less skew than the original FES (-2.4 and -3.3, respectively). Factor analysis of the MFES revealed two factors accounting for 75% of the sample variance, grouping into an 'indoor type activity' factor and an 'outdoor type activity' factor. Retest reliability for the MFES was high (intraclass correlation coefficients = .93). Significant differences were evident between the FBC group and the healthy older group on all items of the MFES and on the total MFES score (p < .05). On the basis of these preliminary findings, the MFES appears to be a reliable and valid measure of falls self-efficacy, and could be a useful addition in the comprehensive assessment of older people with balance disturbance or falls.