Field observation, field cages, and laboratory arenas were compared as methods to estimate daily per capita consumption for larvae of Harmonia axyndis Pallas (Coleoptera: Coccinellidae) and Chrysopa nigricornis Burmeister (Neuroptera: Chrysopidae), the two most abundant predators of Hyalopterus pruni Geoffroy (Hemiptera: Aphididae) in prune orchards in California's Central Valley. Daily per capita consumption increased with larval size, and the highest estimates were obtained from the field observation method and the lowest from the laboratory arena method for both predator species. Possible explanations for differences between estimates for each method are explored in detail, the most important of which is the need to measure both biomass killed and biomass consumed, because H. axyridis consumed nearly all of each prey item, whereas C. nigricornis always killed more biomass than they consumed. This study suggests that the laboratory arena method can lead to underestimation of daily consumption and that field cages may be more appropriate for quantifying daily consumption when prey are colonial and predators are relatively immobile. For highly mobile predators or predators of dispersed prey, the field observation method, combined with observations of both the duration and pattern of feeding activity throughout the day, is the best option for quantifying daily per capita consumption.