Long summer break 'has no effect'
But a study involving more than 370,000 North Carolina pupils has found "year-round" schools fare no better than those on traditional calendars.
US schools have even more reason to worry about the "summer learning loss" - now being discussed in the UK - because the break can stretch to three months. However, research by Bradley McMillen, of the North Carolina Department of Public Instruction, raises new questions about the value of re-arranging the academic year.
In North Crolina, many schools now have 15-day breaks every 45 days. But when McMillen compared pupils in 106 "year-round" schools with those in conventional schools he found no statistically significant differences in reading or maths achievement.
Further analysis of test scores suggested that lower-achieving students might benefit slightly from year-round schools. But McMillen says other factors - such as holiday remedial classes offered by year-round schools - could have been responsible.
"A statewide evaluation of academic achievement in year-round schools", by Bradley J McMillen, North Carolina Department of Public Instruction