One Italian 30 cum laude, three possible French grades

An exchange semester ends with a transcript in someone else's grades. A 27 out of 30 says little to a French department where fewer than 7 percent of passing grades are above 14. The European Commission's method is not an exchange rate between scales but a comparison of positions: where a grade sits among its own department's passing grades, and which grade sits in the same place on the other side.

What a grading table is

The ECTS Users' Guide 2015 asks each institution to publish, for each field of study, the share of passing grades that went to each grade over at least the last two years, with fails left out (sections 4.3 and 4.4). Read the shares from the best grade down, add them as you go, and every grade gets a band. On the engineering table the guide prints for FH Vorarlberg in Austria, a 1 went to 34 percent of passing results, a 2 to 37, a 3 to 19 and a 4 to 10. A 2 therefore covers the band from 34 to 71 percent, counted from the top.

The Commission's own example

Annex 2 sets that table beside Ghent's engineering table, which runs from 20 down to 10. Ghent's 14 covers 29.73 to 43.16 percent, its 13 covers 43.16 to 57.51 and its 12 covers 57.51 to 72.31, so all three overlap FH Vorarlberg's 2. The guide says the ranges typically overlap and that the receiving institution should decide in advance whether it takes the minimum, average or maximum comparable grade: here 12, 13 or 14. Both institutions in the example chose the average, and the 2 becomes a 13.

The return trip is cleaner. A student coming home from Ghent with an 11 holds a grade covering 72.31 to 85.39 percent, which sits entirely inside FH Vorarlberg's 3 (71 to 90). Every rule gives a 3, and so does the guide.

One 30 cum laude, three answers

The rule matters most near the top of a scale, where bands are thin. In 2015 the Egracons project, the EU co-funded conversion tool, worked an Italian example table against the Université de Rouen's. On the Italian table a 30 cum laude went to 5.7 percent of passing grades. At Rouen that top 5.7 percent overlaps five grades: 15 (3.01 to 6.66 percent), 16, 17, 18 and 19 (the top 0.06). The minimum rule gives 15, the average, read as the middle overlapping grade, gives 17, and the maximum gives 19.

The project's own documents do not agree on which to use. Its grading table principle page converts the 30 cum laude to a 15, "obtained by 6.66%, which is closest to 5.7%". Its introduction to grading tables, dated 9 March 2015, says that where several grades overlap "the median is taken", which is 17. As of September 15, 2026, this page holds both readings and chooses neither, and the calculator prints all three.

Lower down the scale the rules agree more often. The Italian 27 covers 34.2 to 46.0 percent and Rouen's 12 covers 28.17 to 46.62, so the 27 is a 12 under every rule, the conversion the introduction reports. The principle page adds that a grade below 25 becomes 10. From 18 to 23 that holds under every rule, but the 24 covers 63.2 to 74.5 percent and overlaps Rouen's 11 for 2.01 points before the 10 begins, so it is a 10 only under the minimum rule and an 11 under the maximum.

When two grades share the middle

The Users' Guide never defines "average", and an even number of overlaps has two middle grades. Its other Annex 2 example settles one case: a Paris grade whose band overlaps Rome's 27 and 28 becomes a 27 under the minimum or the average rule and a 28 under the maximum. The calculator follows it and names both middle grades.

The retired letters are still printed

Before these tables, ECTS used fixed letters: A for the best 10 percent of passing grades, B the next 25, C the next 30, D the next 25 and E the last 10. The 2015 guide says those scales "are not used anymore". The University of Padova still prints them beside its 2025/26 tables for partners that send only a letter, and its page works one: a Spanish partner's transcript carrying an ECTS C, which on the Astronomy table corresponds to marks from 24 to 28. The overlap method reproduces that range from the current table. A C covers 35 to 65 percent; Padova's 28 covers 29.24 to 39.38 and its 24 covers 64.05 to 71.16. Padova also prints a grade that straddles a letter boundary with both letters, so its 30 is A-B and its 24 is C-D, and all fourteen Astronomy labels match the overlaps.

Pavia letters its tables another way. Its 2024 engineering table gives each grade one letter, taken from the share at or above that grade, and under the calculation rule it published for 2024 no letter may be skipped, so its 30 cum laude is an A although it covers 12.01 percent. A 27 there covers 30.44 to 38.40 percent and is printed C, where Padova's convention would print B-C. Four of Pavia's fourteen engineering grades change label.

What the arithmetic cannot settle

A conversion is only as fair as the pairing: the guide asks for comparable reference groups in the same field of study, built from enough results to be reliable. The calculator converts whatever two tables it is given and cannot check either.