This paper reports an experimental evaluation of a direct instruction procedure in which children are taught and practise activities to both foster and monitor their comprehension while reading. Three different class groups of low comprehenders, Standard 4, Form 1 and Form 2, were serially exposed to 21 days of 20–25 minutes instruction in four study activities: summarising, questioning, clarifying and predicting. Instruction was by reciprocal teaching whereby the adult tutor and children took turns leading a dialogue centred around the important topics of the instructional text. The design of this study incorporated both with-subjects multiple baseline and between-subjects comparisons, the no treatment comparison groups being average and above average comprehending class peers. Daily comprehension assessment on a different text at the same difficulty level as that used for reciprocal teaching and pre and post intervention scores on a reading comprehension test served as the dependent measures for the study. Results show significant increases in accuracy on comprehension tests for all experimental groups upon introduction of the reciprocal teaching procedure. On completion of the intervention two of these groups scores matched those of the above average controls. Follow-up probes eight weeks later showed that all groups had maintained their comprehension gains.