Teacher scaffolding and more specifically, support that is adapted to or contingent upon a student's understanding, is considered effective in promoting student learning. Increasing control upon a student's failures, decreasing control upon a student's successes and keeping the degree of control the same in the case of partial student understanding are considered contingent support which provides an appropriate level of challenge. However, these theoretical premises have not yet been tested in classroom situations. The main goal of this study was to investigate different patterns of contingency and to explore how contingency affects student learning. Twenty-two pre-vocational teachers' lessons were analysed regarding contingency and student learning, using micro-analysis. The results show that: (1) if the initial student understanding is poor, contingent support results in increased student understanding, and (2) teachers seldom underestimated students' understanding but often overestimated students' understanding. Contingent support can be effective and future research should focus on facilitating teachers in learning how to correctly diagnose students' understanding. (C) 2013 Elsevier Ltd. All rights reserved.