Effective environmental education requires the identification of relevant learning content. Correlational evidence suggests that different environmental knowledge types influence behavior. We explored this experimentally, assigning N = 304 university students to five conditions where they read texts imparting: (1) a moderate amount of system knowledge, (2) a moderate amount of system and action-related knowledge, (3) a moderate amount of system, action-related and effectiveness knowledge, (4) a high amount of system knowledge only, and (5) environment-unrelated knowledge. We then measured participants' intention to consume animal products (operationalized by CO₂ emission score in kg) and their actual consumption behavior (i.e., choosing between vouchers for vegan, cheese, chicken, or beef sandwiches and between vegan or milk chocolate bars). Assessed covariates were participants’ pre-intervention consumption of animal products and affinity for animal products. Results indicate no statistical differences between the different environmental knowledge types on participants’ consumption intentions and actual behavior. However, compared to environment-unrelated knowledge, the intention was lower for system, action-related and effectiveness knowledge (adj. Cohen’s d = 0.52, 95% CI [0.16, 0.87]) and high amount of system knowledge only (adj. Cohen’s d = 0.59, 95% CI [0.23, 0.85]). We discuss the need for further experimental approaches in environmental knowledge research.