This is a time of enormous investment in new healthcare buildings. The UK plans to create upwards of a hundred hospitals and thousands of primary-care clinics and surgery centres. In the USA, more than $200 billion will be spent on new hospitals over the next decade. This wave of construction provides an opportunity to create better buildings by use of the emerging science of evidence-based design, in which the design process is guided by an empirical understanding of the eff ects of health-care physical environments on safety, effi ciency, and clinical outcomes. The scientifi c foundation for evidence-based healthcare design is already large and surprisingly strong. A joint project at Texas AM infection rates are lower when patients are in single rather than multi-bed rooms and air quality is very good. Single rooms have advantages over multi-bed rooms in reducing airborne transmission through ventilation and the possibility of such air quality measures as fi ltration, air changes, creation of positive air pressure to protect an immunocompromised patient from airborne pathogens in adjacent spaces, or creation of negative room pressure to prevent a patient with an aerially spread infection from infecting others. Other evidence suggests that single rooms also help to decrease risk of infections spread by contact. Single rooms are easier than multi-bed rooms to clean thoroughly after a patient leaves, lessening the problem of contaminated environmental surfaces remaining as pathogen reservoirs that can spread infection through hand contact. Single rooms also enable separation or isolation of patients on admission, so that those with unrecognised infections can be identifi ed and assignment to multi-bed rooms with uninfected patients avoided. A major source of contact transmission of infection is unwashed staff hands, and low hand-washing rates among staff are well documented. Evidence suggests that hand-cleaning compliance is worse when sinks and dispensers of alcohol hand gel are positioned several paces away from staff work or movement paths, or behind doors or in other locations outside the visual fi elds of busy nurses and physicians. There are encouraging indications that installation of wash basins and gel dispensers close to staff work paths in visually prominent locations near the point of care leads to sustained increases in hand washing. Floor layouts can be designed to reduce staff fatigue and increase time for care. Conventional fl oor layouts for patient-care units generally provide corridors organised around a central nursing station, where medications and charts are located, and a central location for supplies. Nurses therefore spend much of their time walking up and down halls engaged in wasteful activity, such as fetching supplies, thereby increasing fatigue and shrinking the time available for observing patients and delivering direct care. By contrast, fl oor layouts with decentralised nurse charting or observation stations and supplies dispersed to be close to patients’ rooms cut staff time spent walking and fetching and greatly increase time for observation and care of patients. Properly designed Roger Ulrich is Professor of Architecture at Texas A&M University and a faculty fellow of the Center for Health Systems and Design, housed jointly in the colleges of architecture and medicine at Texas A&M University. He is currently Visiting Professor to the Bartlett School of Architecture, University College London. He has previously served as Senior Adviser on the Environment for Patient Care to the UK National Health Service.