This thesis focuses on developmental errors of Czech learners
of EFL. Data consisting of informal emails with the length of
200-250 words were collected from 15 students at B1 level, and
19 students at B2 level of proficiency. The emails were
transcribed, corrected, and error tagged by using the Louvain
error taxonomy. The errors were divided into 6 domains and 25
categories, which were first analyzed within one proficiency
group, and the findings were then compared. The computer-aided
error analysis showed that the error domains did not change
significantly with a rising proficiency level.