Research about student self-assessment (SSA) has shown that academic performancetends to increase for students trained in SSA, but also that SSA training may increasethe use of self-regulated learning (SRL) strategies. Consequently, there is widespreadadvocacy for SSA, not least through the “assessment for learning” reform agenda. The educational gains from SSA are suggested to be related to the enhancement ofownership of learning and use of self-regulatory strategies. In addition, SSA has beensuggested to support students’ self-efficacy (SE). These relationships between SSA andSRL/SE, have been claimed theoretically, but without proper empirical support. Thisstudy therefore uses meta-analytic methodology to explore the relationships betweenSSA and SRL/SE. The 21 studies included in the analysis were found by a databasesearch and meta-analyses were conducted using a random-effect model. Inter-studyvariance was estimated by the maximum likelihood method. The findings from thisstudy confirms the theoretical connection between SSA and SRL/SE. By training inSSA, students’ use of self-regulating strategies for learning increase, while their use ofperformance/avoidance strategies decrease. SSA interventions also support students’ SE(.73), meaning that they have a more adequate perception of their own capabilities. Thefindings also confirm what has been noted in some of the individual studies on theeffects on SRL/SE from SSA interventions that girls tend to benefit more from suchinterventions, at least in relation to SE.