To establish the relative clinical effectiveness and cost-effectiveness of paracetamol plus ibuprofen compared with paracetamol and ibuprofen separately for time without fever, and the relief of fever-associated discomfort in young children who can be managed at home. The trial design was a single-centre (multisite), individually randomised, blinded, three-arm trial comparing paracetamol and ibuprofen together with paracetamol or ibuprofen separately. There were three recruitment settings, as follows: 'local' where research nurses were recruited from NHS primary care sites; 'remote' where NHS sites notified the study of potentially eligible children; and 'community' where parents contacted the study in response to local media advertisements. Children aged between 6 months and 6 years with fever > or = 37.8 degrees C and < or = 41 degrees C due to an illness that could be managed at home. The intervention was the provision of, and advice to give, the medicines for up to 48 hours: paracetamol every 4-6 hours (maximum of four doses in 24 hours) and ibuprofen every 6-8 hours (maximum of three doses in 24 hours). Every parent received two bottles, with at least one containing an active medicine. Parents, research nurses and investigators were blinded to treatment allocation by the use of identically matched placebo medicines. The dose of medicine was determined by the child's weight: paracetamol 15 mg/kg and ibuprofen 10 mg/kg per dose. For additional time without fever in the first 4 hours, use of both medicines was superior to use of paracetamol alone [adjusted difference 55 minutes, 95% confidence interval (CI) 33 to 77 minutes; p < 0.001] and may have been as good as ibuprofen (adjusted difference 16 minutes, 95% CI -6 to 39 minutes; p = 0.2). Both medicines together cleared the fever 23 minutes (95% CI 2-45 minutes; p = 0.015) faster than paracetamol alone, but no faster than ibuprofen alone (adjusted difference -3 minutes, 95% CI 24-18 minutes; p = 0.8). For additional time without fever in the first 24 hours, both medicines were superior to paracetamol (adjusted difference 4.4 hours, 95% CI 2.4-6.3 hours; p < 0.001) or ibuprofen (adjusted difference 2.5 hours, 95% CI 0.6-4.5 hours; p = 0.008) alone. No reduction in discomfort or other fever-associated symptoms was found, although power was low for these outcomes. An exploratory analysis showed that children with higher discomfort levels had higher mean temperatures. No difference in adverse effects was observed between treatment groups. The recommended maximum number of doses of paracetamol and ibuprofen in 24 hours was exceeded in 8% and 11% of children respectively. Over the 5-day study period, paracetamol and ibuprofen together was the cheapest option for the NHS due to the lower use of health-care services:14 pounds [standard deviation (SD) 23 pounds] versus 20 pounds (SD 38 pounds) for paracetamol and 18 pounds (SD 40 pounds) for ibuprofen. Both medicines were also cheapest for parents because the lower use of health care services resulted in personal saving on travel costs and less time off work: 24 pounds (SD 46 pounds) versus 26 pounds (SD 63 pounds) for paracetamol and 30 pounds (SD 91 pounds) for ibuprofen. This more than compensated for the extra cost of medication. However, statistical evidence for these differences was weak due to lack of power. Overall, a quarter of children were 'back to normal' by 48 hours and one-third by day 5. Five (3%) children were admitted to hospital, two with pneumonia, two with bronchiolitis and one with a severe, but unidentified 'viral illness'. Young children who are unwell with fever should be treated with ibuprofen first, but the relative risks (inadvertently exceeding the maximum recommended dose) and benefits (extra 2.5 hours without fever) of using paracetamol plus ibuprofen over 24 hours should be considered. However, if two medicines are used, it is recommended that all dose times are carefully recorded to avoid accidentally exceeding the maximum recommended dose. Manufacturers should consider supplying blank charts for this purpose. Use of both medicines should not be discouraged on the basis of cost to either parents or the NHS. Parents and clinicians should be aware that fever is a relatively short-lived symptom, but may have more serious prognostic implications than the other common symptom presentations of childhood.