Abstract The effectiveness of different types of written corrective feedback has been investigated over the last twenty years but it is still not possible to make firm conclusions about which options are the most beneficial to ESL learners. This article first provides an overview of the currently available research findings and then presents the results of a six month study of the relative effectiveness of providing thirty-nine low intermediate ESL learners in Auckland, New Zealand, with three different direct written corrective feedback options. Assigned to three groups (direct corrective feedback, written and oral meta-linguistic explanation; direct corrective feedback and written meta-linguistic explanation; direct corrective feedback only), the subjects produced four pieces of writing (pre-test, immediate post-test, and two delayed post-tests). Two functional uses of the English article system (referential indefinite “a” and referential definite “the”) were targeted in the feedback. No difference in effect upon accuracy was found between the three treatment options, suggesting that the provision of error correction alone may be sufficient for learners at a low intermediate proficiency level.