The aim of this article is to describe and compare the results of written examinations in mathematics at the University of Finance and Administration in Prague, which were conducted in three different forms: standard (face-to-face), correspondence (individual test preparation sent electronically for checking) and online (under the direct supervision of an examiner via camera). The main objective is to determine whether the two alternative forms (correspondence or online) are suitable substitutes for the standard form. For a better comparison, the results of the summer semesters of three consecutive academic years, when a different form of examination was used each time, are used. Statistical evaluation showed that standard and correspondence forms of testing led to similar student performance, while students were less successful online. Possible reasons for the results are analysed.