In this paper I would like to give an overview of my several years' experience 
in exploring PICLE - the Polish subcorpus of the International Corpus of Learner 
English, which began its existence in 1995. A brief introduction to current 
computer learner corpus research will be offered, followed by a presentation of 
a contrastive research scheme, in which findings from PICLE are juxtaposed with 
data derived from a series of other, comparable corpora of native English, 
learner English, and native Polish. The main part of the paper will be taken up 
by a presentation of selected examples of findings obtained to date, taken from 
the areas of lexis, phraseology and stylistics. It is hoped that the discussion 
will convincingly underscore the big potential of learner corpora as a source of 
knowledge about Polish students' language performance and as a resource for 
tuning in task-oriented language courses in Poland.