A new exam "beyond A-level" is needed to stretch the most able in maths, the exams regulator has said in a report which shows that most teachers believe the present qualification has got easier.
Six out of 10 teachers said the exam, which was redesigned in 2004, is not suitable for the brightest students. The report said it failed to increase the participation of less able pupils in the subject.
The Qualifications and Curriculum Authority said its investigation into the new AS- and A-level in maths did not conclude whether the exam is easier.
The regulator said the relationship between maths and further maths should be severed: some modules can count for both qualifications. Doug French, spokesman for the Mathematical Association, said teachers were happier with the new maths A-level but it was hard to say whether it was actually easier.
"It is very difficult to make a clear judgment when you are not comparing like with like," he said.
The new A-level was introduced in response to a crisis following the introduction of Curriculum 2000. Pass rates and pupil numbers fell as students were expected to do more in the first year of the subject.
The changes allowed students to make up the six modules needed for the qualification from two A2 and two AS modules with a further two optional modules - which could be at AS-level.
They found that the majority of teachers believe that the transition between GCSE and A-level is now easier.
The changes have led to an increase in the take-up of maths at AS- and A2-level, with greater rates of retention from AS to A2.
A large number of pupils resat AS units, with 40 per cent of those surveyed saying "all students" would take resits. But although more pupils are now taking A-level maths, they still tend to be recruited from a "clever core" of students.
The report said that while many pupils rate maths as the most difficult subject they study, large numbers are able to pass A-level maths and further maths in only slightly more time than taking one A-level in another subject.
The report is based on national examinations data, a large-scale questionnaire and a sample of 19 schools, speaking to both staff and students.