Background Unintentional falls from windows are an important cause of paediatric morbidity. There have been no controlled studies to identify modifiable environmental risk factors for window falls in young children. The authors have piloted a case–control study to test procedures for case identification, subject enrolment, and environmental data collection.

Methods Case windows were identified when a child 0–9 years old presented for care after a fall from that window. Control windows were identified (1) from the child's home and (2) from the home of an age- and gender-matched child seeking care for an injury diagnosis not related to a window fall. Study staff visited enrolled homes to collect window measurements and conduct window screen performance tests.

Results The authors enrolled and collected data on 18 case windows, 18 in-home controls, and 14 matched community controls. Six potential community controls were contacted for every one enrolled. Families who completed the home visit viewed study procedures positively. Case windows were more likely than community controls to be horizontal sliders (100% vs 50%), to have deeper sills (6.28 vs 4.31 inches), to be higher above the exterior surface (183 vs 82 inches), and to have screens that failed below a threshold derived from the static pressure of a 3-year-old leaning against the mesh (60.0% vs 16.7%). Case windows varied very little from in-home controls.

Discussion Case–control methodology can be used to study risk factors for paediatric falls from windows. Recruitment of community controls is challenging but essential, because in-home controls tend to be over-matched on important variables. A home visit allows direct measurement of window type, height, sill depth, and screen performance. These variables should all be investigated in subsequent, larger studies covering major housing markets.