This study focused on the beliefs of teachers that comprised 'teacher efficacy' and explored meanings of the construct of teacher efficacy in terms of various cognitive motivational constructs and teacher characteristics contended as predictive of teacher efficacy. Urdu version of the Teacher Efficacy Scale (TES) and a set of indigenous questionnaires; Task Motivation Style (TMS), Beliefs about Intelligence as Incremental Quality (BIIQ), and Ability-Effort attribution matrix were administered to 227 mostly in-service teachers. Confirmatory factor analysis upheld teaching efficacy (TE) and personal efficacy (PE) as discrete factors operative in TES similar to the findings of several Western studies. Pearson correlations between the motivation constructs/measures; TMS, BIIQ, Ability-Effort attribution, and TE/PE aspects of efficacy supported the construct validity of TES. predictive model conceptualized with BIIQ and TMS as core exogenous variables, Ability-Effort attribution as context/intervening variable and TE/PE as effect variable was tested through multiple regression and path analysis techniques. Task motivation (B2 = .623, p<.00l), and incremental concept of intelligence (B1 = .252, p<.05), strongly predicted TE (p<.01). However, PE could be predicted less significantly (p<.05) and in other than the expected pattern by the same variables. A perception of intelligence as a fixed non-incremental entity and ability attribution as causal of student achievement were found to underlie PE depicting it to be a rather weak motivational attitude. However, Anova statistic supported significant interaction effect of TE by PE on task motivation F ( 3,223) = 19.30, p<.00l, incremental concept of intelligence F(3,223) = 2.84, p<.05, ability attribution for failure, F (3,223) = 9.95 p<.0l, and alternately effort attribution for success F (3,223) = 2.56, p<.004. Under TE by PE interactive conditions, high PE, high TE teachers' mean rating for students' 'ability' as causal of their success and 'effort' as causal of their failure was found to be significantly lower, and in reverse direction than that of low (10 PE, 10 TE and 10 PE, hi TE) efficacy groups of teachers respectively. Thus high TE by PE conditions/profile reflected motivationally adaptive outlook (in terms of their beliefs in intelligence to increase with effort and salience of task-orientation in teaching) as teacher; low TE by low PE profile teachers were deficient in these characteristics. We found a high PE attitude in the presence of high TE conditions as potentially motivating the teachers, but high PE under low TE conditions as less motivating. In this context of profile analysis, studies that combine PE and TE score may be misleading. This study recommends as interaction application of PE by TE dimensions since the two co-responde. Antecedent teacher characteristics; gender and experience were not found to be related to teacher efficacy, however Science vs. Arts teacher dichotomy was related to teacher efficacy scores (rpbi = .151, t = 2.31, p<.05). Science teachers had higher TE score than Arts teachers. The findings bear important implications for teacher education directly and for school education indirectly in Pakistan. In this context it is suggested that belief re-cognition intervention as a part of teacher education programme (as internal motivation force), and a target teacher-performance reinforcement contingency (an external reason) may be applied as a two-pronged strategy to mediate teaching practices in Pakistan for raising student outcome and rationalizing teachers' sense of efficacy. Beliefs as 'personal-knowledge' together with 'professional-pedagogical knowledge' drive classroom actions. Prospective teachers thus may preferably be screened for their beliefs and frames of mind or/and their training must catter to aligning and fostering such beliefs and cognitions as support their teaching intervention with a goal and effect.