Aims and objectives: English has become the dominant donor language for many languages, including Croatian. Perception of English loanwords has mainly been investigated through corpus-based studies or attitude questionnaires. At the same time, normative data for unadapted English loanwords are still mainly unavailable. This study aims to fill that gap by collecting affective and lexico-semantic norms for unadapted English loanwords in Croatian. Methodology: Valence, arousal, familiarity, and concreteness ratings for unadapted English loanwords and three types of Croatian equivalents were collected from 565 participants. Data and analysis: Affective and lexico-semantic norms for each word on the four variables are available in the database. In addition, the relationship between different variables was examined. Finally, the differences between English loanwords and three types of Croatian equivalents (in-context, out-of-context, and adapted forms) are reported. Findings: Valence ratings for unadapted English loanwords differed from out-of-context equivalents and adapted forms. Unadapted English loanwords were rated as more arousing than Croatian equivalents. Finally, unadapted English loanwords were less familiar and less concrete than in-context and out-of-context equivalents. The findings suggest that Croatian speakers perceive unadapted English loanwords differently on affective and lexico-semantic levels compared with Croatian equivalents. Originality: This is the first study to provide affective and lexical norms for 391 most frequent unadapted English loanwords in Croatian. Implications: The reported normative data will contribute to the existing knowledge about the processing of English loanwords by enabling experimental research on this topic.