The understanding that learning develops best in participatory, collaborative, interactive partnerships suggests that attempts to establish such relationships in classrooms deserve scrutiny. This study investigates co-constructive approaches to teaching and learning in a New Zealand secondary school. It studies co-construction from the perspectives of seven mathematics teachers and one art teacher and their students, as the teachers attempted to co-construct aspects of the classroom curriculum with students. This study focused on the practices, understandings and expectations of both teachers and students when co-constructing. In addition, it identified the potential constraints and difficulties that impacted on such practices in this secondary setting. Ideally, such classrooms demonstrate collaborative, democratic, socially just practices with sharing of power between teacher and student: ideas emanating from theorists such as Dewey, Bruner, Freire and Vygotsky. This interpretive study responds to the dearth of research in New Zealand secondary classrooms about such approaches. It draws on principles and methods of grounded theory (Charmaz, 2003, 2006) to analyse the rich data generated by multiple methods. This allows the meanings of participants to prevail, by guarding against the imposition of meanings from either literature or researcher. Methods of data collection included classroom observation, interviews with teachers and groups of students, audiorecording and various forms of documentation sourced from these. The findings showed co-constructive practice has shared characteristics, such as spontaneity, quality teacher-student communication with a stress on the teacher as a listener, active participation by learners, and a distinction of varying qualities of practice. However, classrooms are not formulaic, as each teacher evolves their own approach. Initially, strong subject pedagogical traditions and mythologies emerged as barriers to co-construction, particularly in mathematics. Learning is involved for all participants in pedagogical change. Students generally responded positively as co-construction provided opportunities for some input, choice and control of their learning which improved relevance, motivation and responsibility. While very aware of the issues related to the use of particular strategies such as groups, the positives of helping each other learn prevailed. The numerous potential constraints to easy implementation of co-construction reflected the diversity of teacher and student personal traits and beliefs about pedagogy, roles, and responsibilities. Moreover, teacher subject and pedagogical knowledge had import, as did their need to establish control and a relationship of trust with a class. The difficulty of secondary settings was endorsed. However, despite these, aspirational reasons for continuing to co-construct were expressed in the benefits experienced by both teachers and students. This study demonstrates how co-construction features in the daily interactions which comprise the classroom curriculum, even where there are constraints to formal consultation over content of classroom programmes. If the bi-directionality of classroom interactions is acknowledged, students have spaces for influence, input and control. It affirms the importance and the diversity of perspective that both teachers and learners bring to the partnership. The need to persevere to establish and grow this approach in secondary schools is strongly advocated. This approach maximises respectful relationships, dialogue, innovation, and excitement in teaching and learning.