Service-Learning programs at school and at university level are commonly considered as reliable instruments to increase civic engagement. In the last years, within Service-Learning Research a new paradigm arose understanding itself as non-foundational and critical. This study differentiates this critical perspective by offering a secondary analysis of available evaluations of Service Learning in the empirical and normative dimension. The analysis reveals that the empirical impact of Service-Learning instruments on political participation is not as distinct and clear as commonly expected. Evaluation results are often inconsistent and make general positive statements difficult. The normative evaluation shows that one has to avoid Service-Learning programs generating depoliticized participation and to impede tendencies to misuse Service-Learning programs as a mere substitute of governmental social policy. All things considered turning Service Learning into an effective means to promote participation, the existing programs have to be improved.