The article discusses an experiment that looked into the acquisition of collocational knowledge in three university groups studying online, each subjected to different learning conditions: incidental acquisition, intentional acquisition, and intentional acquisition with an extra productive output (essay), the latter having been assessed for the amount and accuracy of target lexis usage in their texts. The aim of the study was to see how well upper-intermediate university students could identify collocations in an input text, and how the text-based output affected the collocational uptake outcomes. The study showed that the productive output group outperformed the other intentional learning group, while incidental acquisition group failed to complete a productive knowledge posttest. Although the study revealed only slightly higher gains in the output group, their results appeared more consistent than those demonstrated by the other intentional uptake group, whose retention rate decreased by the time of delayed posttest.