Prosocial behavior is widely recognized for its benefits, yet its underlying mechanisms remain insufficiently understood, particularly in ecological contexts where effort is required. This thesis advances this understanding through three studies grounded in appraisal theories: two empirical studies and one theoretical proposal. The empirical studies introduced an effort-based paradigm, the slider task, requiring participants to complete time-limited motor tasks in dyadic (Study 1) and multi-agent (Study 2) settings. Performance metrics, perceived importance, emotional reward, and individual differences in prosocial preferences were collected to examine affective dynamics. Results supported the hypotheses that prosocial action induces affective stress and that emotional rewards are modulated by different configurations of proself and prosocial benefits. Follow-up analyses revealed enhanced performance under high-stakes conditions and a dominance of proself bias in subjective satisfaction. Study 3 addressed help-seeking behavior, proposing an integrative framework that combines interpersonal emotion regulation, learning, and affective prediction across three phases: contemplation, execution, and resolution. Together, these studies conceptualize prosocial behavior as a dynamic, effortful, and inherently interpersonal process.