PurposeA number of empirical studies and evaluations in Norway and Sweden shows variabilities in the degree to which the municipalities succeed in their endeavors to support school principals’ instructional leadership practices. In response to this situation, the Norwegian and Swedish directorates of education have developed a joint collaborative design for practice learning of instructional leadership. Based on findings from two separate studies, the purpose of this paper is to contribute to theory development and improved practice for school district administrators and their subordinated school leaders.Design/methodology/approachThe study draws on the data from participants who completed the program in June 2015, June 2016 and June 2017, respectively. The data are based on individual reflection documents from students on their learning and new leadership practices 4 months, 16 months and 28 months after the end of the program.FindingsThe project subjected to this study, labeled “Benchlearning,” involved learning from experiences of others, observational learning, dialogic group learning and in the final round translating what is learnt into the social and cultural context in which the individual school principal’s school is situated. When participating school principals experience observation-based learning together with trusted colleagues, followed by vicarious learning from these experiences in their schools, the authors see some facilitating factors to be of particular importance: learning infrastructure, digital tools, compulsory tasks associated with preparation and subsequent experiments with their teachers. Emerging from the analysis was a systematic balancing act of autonomy and structure running through the various learning activities. Finally, a strong evidence was found that developing core competence in digital learning and formative assessment among teaching staff required enhanced distributed leadership across the whole school organization. By sharing leadership tasks on instructional issues with teachers and other non-leaders, principals succeeded in leveling up instructional leadership significantly.Research limitations/implicationsThe implications of the study can be summed up in the following four principles. First, policy makers should take into accounts the fact that principals’ motivation and willingness to initiate change processes can be created in a synergy between structured school visits and engagement in learning groups based on a sound theoretical foundation. Second, within a socially contracted practice in a well-designed learning group, it is possible to raise principals’ level of self-efficacy. Third, a systematic reflection process on authentic practice is an example of how principals can develop their metacognitive capacity and how knowledge can be transformed into new practice. Finally, educators should be trained to be process leaders in order to create a balance between demand and support in promoting principals’ learning of new instructional leadership practices.Practical implicationsSchool district administrators should take into accounts the fact that changing practices will be supported by sense-making processes involving discussions about how new instructional practices are justified. Specifically, shifts in talk and actions will also involve shifts in the ways people relate to each other and how they relate to their internal context. Further, leadership programs should include trying out new practices as the focal learning mode, accompanied by individual and collective reflective activities.Originality/valueThe findings of the study underscore the mutual interdependence of distributed leadership and student-centered focus accompanied with the school’s learning capacity as enabling conditions for principals’ practice learning in the field of instructional leadership.