This thesis examines the classroom emotions of sixty-one 9 - 11 year old elementary school children in two classrooms as they engaged in learning tasks. Emotion data were obtained within the larger context of the Project on Learning using a multi-modal methodology. Comprehensive data collection included continuous observations with concurrent video- and audio-recorded data also collected over two full units of learning in science and social studies. Video-cued interviews and a modified experience sampling methodology (ESM) obtained retrospective and on-line emotion data respectively. Statistical analyses of the ESM data for the whole sample of sixty-one children identified three significant emotion factors - Factor 1 (negative), Factor 2 (positive) Factor 3 (happy), which were then tested against gender, achievement and concept learning. Indepth qualitative analyses of eight specific target children selected from the wider sample provided the contextual and within-child variables in relation to the significant emotions obtained. Specific antecedents were identified for events where children reported some or all of the significant emotion factors while selecting and/or engaging in learning tasks. Findings include a significant gender difference with girls reporting positive emotions more frequently than boys, and that irrespective of individual differences in their achievement measures, when new information was learned the target children reported feeling any or all of the Factor 2 emotions. This finding at a qualitative level was contradictory to the aggregated finding that there was no significant relationship between concept learning and Factor 2. This inconsistency was consistent with previous literature comparing and arguing for both nomothetic and idiopathic methods in emotion research.