This study aims to compare distance learning innovations carried out by Islamic Religious Education teachers in madrasahs and schools. The innovations compared are based on five aspects of Roger's theory, namely: relative advantage, compatibility, testability, observability, and complexity. The research method used is a comparative study with a quantitative approach. Data was collected using a Likert Scale questionnaire that used four levels of choice. The data obtained from the field were analyzed objectively. The data analysis technique used Mann-Whitney U as a non-parametric test. This analysis technique was chosen because the data were not normally distributed. The results of the hypothesis test show that there is no difference in innovation in terms of relative advantage, compatibility, testability, and observability because of the Asymp. Sig 2 tailed score is greater than 0.05. Asymp. Sig score on relative advantage 0.768. Asymp. Sig scores on compatibility 0.170. Asymp. Sig scores on testability 0.215. Score on observability 0.173. The difference in innovation is only in the aspect of complexity with a score of Asymp. Sig 2 tailed 0.007< 0.05. The study found that the innovations carried out by the two groups of Islamic Religious Education teachers were relatively the same, only differing in one aspect, namely complexity.