This is a critical analysis of how cultivation has been conceptualized in theory and research. Cultivation indicators are examined for their meaning in texts, the meaning received by viewers, and the distinction between estimations and beliefs. The construct of television exposure is analyzed in terms of the assumptions of uniform messages and nonselective viewing, as well as the conception of time and dominance. The nature of relationship is illuminated through the assumptions of linearity, asymmetry, control variables, causation, level of generality, mainstreaming, and resonance. Recommendations are provided to suggest revisions for conceptualizing the existing theory and extending it. Suggestions for extension include reconceptualizing the effect and the relationship, developing a typology of effects, considering the context of other simultaneous influences, providing analysis over time, and examining the process of influence on individuals and on the messages.