Recycling is an environmentally responsible behaviour that has gained increasing attention and formal implementation over the past two decades. Conservation psychologists interested in understanding proenvironmental behaviour have often turned to the study of recycling as a basis for their investigations. Researchers concerned with the antecedents of recycling have focused on both situational variables (e.g., norms, convenience) and personal variables (e.g., identity, attitudes). However, conclusions about proenvironmental behaviours, such as recycling, necessarily rest on the measurement of such behaviours. Within conservation psychology, self-report measures are often used to determine an individual’s level of involvement in environmentally responsible behaviors. To be sure that such measures are valid, it is necessary to compare data from self-report methods with data from observational methods. The present study assessed the validity of a self-report measure of recycling behavior, and the usefulness of two different personal variables as predictors of recycling behavior. A large sample of Canadian university students (N = 1283) completed a measure of pro-environmental orientation (i.e., New Ecological Paradigm Scale) and a self-reported measure of recycling frequency. From this sample, twenty-two participants completed a personality measure, and a task that was seemingly concerned with their mathematic ability, but that actually measured their recycling behavior. Participants who reported higher levels of recycling behavior recycled their material from the experiment significantly more often than participants who reported lower levels of recycling behavior. Therefore in this study, the self-report and observational measures of recycling displayed convergent validity. As for predictors of recycling behavior, the personality trait of conscientiousness was more positively associated with recycling than one’s pro-environmental orientation as measured by the NEP Scale. Furthermore, a factor analysis of the NEP Scale replicated the dimensional breakdown originally reported by Dunlap. Results from this study are discussed in terms of the need for measurement validity and the need to consider additional personal variables when explaining environmentally responsible behaviours.