This paper identifies, describes and analyzes the development phases of the On-line Mentoring Program (OMP) at the Teachers Portal of Universidade Federal de São Carlos (www.portaldosprofessores.ufscar.br) in view of its participant novice teachers’ demands and their mentors’ subsequent support. This analysis may assist in planning novice teachers’ career development in association with experienced teachers (mentors). Its theoretical framework includes literature on teacher learning and professional development (Shulman 1986, 1987, Ball & Cohen 1999, Borko 2004), the first years of teaching (Grossman, Thompson & Valencia 2001, Zanten & Grospiron 2001), mentoring processes (Weiss & Weiss 1999, Wang & Odell 2002) and distance education (Marcelo Garcia 2002). The mentoring program in question is an on-line program aimed at elementary school teachers at the first grades with up to 5 years of experience. It is carried out by ten mentors coaching 42 novice teachers during 6 to 28 months.

The research design adopted in this investigation can be characterized as action-research founded on constructive-collaborative strategies (Cole & Knowles 1993). Its chief sources of data were: written communications between mentors and novice teachers, mentors’ and novice teachers’ reflexive journals, and reports on observations at weekly meetings between mentors and researchers. Its investigation steps, data collection tools and analysis were defined by the researchers alone whereas the OMP activities were defined by the researchers and mentors. The intervention was constructed jointly. This article examines the mentoring program from a process standpoint considering the contents dealt with by mentors and novices and their communications, such as the latter’s reflexive journals and teaching cases reported by them about the mentoring process and the former’s accounts taken down during the aforementioned weekly meetings. The analysis focuses on the themes worked by the mentors and novices and the support evidenced throughout the mentoring process.

Results indicate that the investigated mentoring processes occur consistent with distinct phases and their intrinsic characteristics, goals and procedures. Each one of the phases (i.e., initial or bonding phase, development or intensification phase, and disengagement phase) implied discussion between mentors and novice teachers about different themes; demanded specific support from mentors in view of novice’s formative needs; and presented critical moments in which clear changes were observed among participants: new stances and learning. The comparison between processes carried out by different mentor-novice dyads indicates that the aforementioned phases were influenced by the novices’ time during the school year. Problems encountered when choosing classes and managing classrooms and lack of knowledge about the school’s culture arise more often in the beginning of the school year whereas difficulties associated with developing and keeping an optimistic disposition towards teaching activities and roles and meeting students’ needs are more frequent in the second semester. Results also point to the seasonality of lived processes (themes worked and support demanded by participants). There is evidence that mentoring processes run more smoothly in the first semester of the school year owing to novices’ more available time to respond to mentors’ demands and suggestions. The second semester seems to be ‘heavier’ and harder for novices to adopt new stances and do tasks proposed by mentors. These characteristics should be taken into account when proposing and carrying out a mentoring program.