To assess the effect of on-going and of previous or near by hunting pressure on game birds in a neotropical rain forest (French Guiana), I compared species abundances between six hunted and disturbed areas, seven non-hunted and pristine areas and eight intermediate areas, undisturbed and not currently hunted, but formerly hunted or close to current hunting areas. I recorded all birds detected within 100 m-wide strip transects, walked at random through every forest type all day long. The frequency of records per 10 h and flock sizes on each transect were averaged over all surveys in each study site (N = 3025 h). Censused groups included terrestrial (tinamous, woodquails, curassow and trumpeter) and arboreal species (guans, macaws, parrots, toucans), but not waterbirds. The abundance of all game birds was higher in non-hunted than in hunted areas, though not always significantly for secretive understorey species (tinamous, guans, wood-quail) or canopy frugivores (parrots, macaws, toucans). The Black Curassow and Gray-winged Trumpeter exhibited the highest and most consistent increases (7–10-fold) and more so between intermediate and non-hunted sites than between hunted and intermediate sites. The mean flock size also increased along the hunting gradient, especially in flocking species (macaws, parrots, trumpeter). Low reproductive rates and/or seasonal movements may explain that hunting pressure could still be felt after hunting has ceased or when it persists only away from a given area. Such a persistent effect would affect some populations in small protected areas.