The standard approach for norovirus control in hospitals in the UK, as outlined by the Health Protection Agency guidance and implemented previously by Lancashire Teaching Hospitals, involves the early closure of affected wards. However, this has a major impact on bed-days lost and cancelled admissions. In 2008, a new strategy was introduced in the study hospital, key elements of which included closure of affected ward bays (rather than wards), installation of bay doors, enhanced cleaning, a rapid in-house molecular test and an enlarged infection control team. The impact of these changes was assessed by comparing two norovirus seasons (2007-08 and 2009-10) before and after implementation of the new strategy, expressing the contrast between seasons as a ratio (r) of expected counts in the two seasons. There was a significant decrease in the ratio of confirmed hospital outbreaks to community outbreaks (r = 0.317, P = 0.025), the number of days of restricted admissions on hospital wards per outbreak (r = 0.742, P = 0.041), and the number of hospital bed-days lost per outbreak (r = 0.344, P < 0.001). However, there was no significant change in the number of patients affected per hospital outbreak (r = 1.080, P = 0.517), or the number of hospital staff affected per outbreak (r = 0.651, P = 0.105). Closure of entire wards during norovirus outbreaks is not always necessary. The changes implemented at the study hospital resulted in a significant reduction in the number of bed-days lost per outbreak, and this, together with a reduction in outbreak frequency, resulted in considerable cost savings. (C) 2011 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.