Eating while watching TV has generally been found to increase both immediate and delayed energy intake. Here we examine two factors – gender and habitual processed-food intake – that may moderate these effects. Participants [n = 153; 95 women, 58 men; Mage = 19.7 (SD = 2.9); MBMI = 22.4 (SD = 3.1)] ate an ad libitum snack either with or without TV, followed around 1 h later by lunch. There was an interaction between TV and gender for both meals. Women tended to consume more snack food in the TV condition, with men consuming more in the no-TV condition. Participants who habitually consumed more processed food also ate more snacks, independent of any other variable, including rated liking. At lunch, men who had earlier snacked with TV ate more than men who had snacked without TV, but this effect was not evident in women. On memory recall, all participants underestimated how much snack food they had eaten, and this was a function of how much they had actually consumed, with greater error only predicted by greater consumption. The results indicate that the effects of TV on eating can vary with gender and that processed-food history can predict snack food intake. While previous findings suggest memory of prior-intake may be impaired by eating while watching TV, the current results suggest this is not necessarily because of TV per se, but because people sometimes consume more food under such conditions.