This thesis look. at two levels of teacher's role conception, 'ideal' and classroom. Findings supported the premise that there would be variance between these two levels and between role segments which form part of role. A major aim was to explain the variance between role conceptiona. The belief was that attitudes and professionalism together with constraining influences would affect role conception. Observational analysis (school visits and actual teaching) was carried out in a selection of comprehensive schools and some of the data wa~ used in qu.ationnaire formation. The aample of 86 was drawn from three schools from a school division within the Inner London Education Autbority. It was hypoth.sized that certain constraining influences would predominate over others in contributing to the variance between role conceptiona. School organization and student ba.ed constraints were of major importance, thus supporting this hypothesis. Inter-relationship. between variables were examined. Becaus. respondents did not answer acroll the .cale, the sample could DOt be divided on a liberal/conservative dicbotomr a. had been anticipated. Through weighting procedure. new variable. were created: role legments (dependent variable.) and independent variable., such as constraint., influence of alters, altars with .imilar viev., hindered by altera, powerlessness, freedom, employers change/conflict, professionalism and authoritarianism. The latter two variables did not prove to be very significant. However chanae/conflict in relation to role was more prominent than individual change/conflict itama in the earlier analysis of the data. Observational data tends to 8upport the later analysis aDd a180 the view of the researcher tbat teachers vere conservative in the way they responded.