Objective: This study analyses the psychometric properties of the Working Alliance Inventory adapted for guided Internet interventions (WAI-I). Methods: We drew on the data set from a multicenter trial that examined a guided Internet intervention (deprexis) for patients with mild to moderate depression. Two hundred twenty-three patients completed the WAI-I and the Patient Satisfaction Questionnaire (ZUF-8) at posttreatment, and the Attitudes toward Psychological Online-Interventions Questionnaire (APOI) at baseline. We ran confirmatory factor analyses (CFA) testing two- and three-factor solutions and calculated Cronbach's α, item-total correlations, and correlations of the WAI-I with APOI and ZUF-8. Results: The results suggested a two-factor solution, with a very good model fit and evidence of factor independency, adequate internal consistency, and external validity for the complete scale and the sub-scales. Conclusions: The WAI-I showed as a reliable and valid instrument to capture alliance in guided Internet interventions, which might facilitate process-outcome research and treatment development efforts.