A methodological difficulty facing welfare research on nonhuman animals in the zoo is the large number of uncontrolled variables due to variation within and between study sites. Zoo visitors act as uncontrolled variables, with number, density, size, and behavior constantly changing. This is worrisome because previous research linked visitor variables to animal behavioral changes indicative of stress. There are implications for research design: Studies not accounting for visitors' effect on animal welfare risk confounding (visitor) variables distorting their findings. Zoos need methods to measure and minimize effects of visitor behavior and to ensure that there are no hidden variables in research models. This article identifies a previously unreported variable—hourly variation (decrease) in visitor interest—that may impinge on animal welfare and validates a methodology for measuring it. That visitor interest wanes across the course of the day has important implications for animal welfare management; visitor effects on animal welfare are likely to occur, or intensify, during the morning or in earlier visits when visitor interest is greatest. This article discusses this issue and possible solutions to reduce visitor effects on animal well-being.