Personal pornography viewing has been associated with lower sexual satisfaction in both experimental and observational research. The language used to hypothesize this relationship typically suggests that it is frequent viewing, rather than infrequent or only occasional viewing, that is responsible for any adverse effects. When the nature of the relationship between a predictor and a criterion depends on the levels of the predictor, a curvilinear relationship is indicated. Nevertheless, studies have assumed linearity in their analytical approach. Curvilinear relationships will go undetected unless they are specifically tested. This article presents results from a survey of approximately 1,500 U.S. adults. Quadratic analyses indicated a curvilinear relationship between personal pornography viewing and sexual satisfaction in the form of a predominately negative, concave downward curve. The nature of the curvilinearity did not differ as a function of participants' gender, relationship status, or religiosity. But the negative acceleration was slightly more pronounced for men than for women, for people not in a relationship than for people in a relationship, and for religious people than for nonreligious people. For all groups, negative simple slopes were present when viewing reached once a month or more. These results are correlational only. However, if an effects perspective were adopted, it would suggest that consuming pornography less than once a month has little or no impact on satisfaction, that reductions in satisfaction tend to initiate once viewing reaches once a month, and that additional increases in the frequency of viewing lead to disproportionately larger decrements in satisfaction.