AIM: To evaluate whether ≥12-hr shifts are associated with a decrease in resource use, in terms of care hours per patient day and staffing costs per patient day. BACKGROUND: Nurses working long shifts may become less productive and no research has investigated whether potential cost savings are realized. METHOD: A retrospective longitudinal study using routinely collected data from 32 wards within an English hospital across 3 years (1 April 2012-31 March 2015). There were 24,005 ward-days. Hierarchical linear mixed models measured the association between the proportion of ≥12-hr shifts worked on a ward-day, care hours per patient day and staffing costs per patient day. RESULTS: Compared with days with no ≥12-hr shifts, days with between 50% and 75% ≥12-hr shifts had more care hours per patient day and higher costs (estimate for care hours per patient day: 0.32; 95% CI: 0.28-0.36; estimate for staffing costs per patient day: £8.86; 95% CI: 7.59-10.12). CONCLUSIONS: We did not find reductions in total care hours and costs associated with the use of ≥12-hr shifts. The reason why mixed shift patterns are associated with increased cost needs further exploration. IMPLICATIONS FOR NURSING MANAGEMENT: Increases in resource use could result in additional costs or loss of productivity for hospitals. Implementation of long shifts should be questioned.