Ticks were sampled by flagging, collecting from the investigator's clothing (walking samples), trapping with dry-ice bait, and collecting from mammal hosts on Fire Island, NY, U.S.A. The habitat distribution of adult deer ticks, Ixodes dammini, was the same in simultaneous collections from the investigator's clothing and from muslin flags. Walking and flagging samples can both be biased by differences between investigators, so the same person should do comparative samples whenever possible. Walking samples probably give a more accurate estimate than flagging samples of the human risk of encountering ticks. However, ticks (such as immature I. dammini) that seek hosts in leaf litter and ground-level vegetation are poorly sampled by walking collections. These ticks can be sampled by flagging at ground level. Dry-ice-baited tick-traps caught far more lone-star ticks, Amblyomma americanum, than deer ticks, even in areas where deer ticks predominated in flagging samples. In comparisons of tick mobility in the lab, nymphal A. americanum were more mobile than nymphal I. dammini in 84% of the trials. Therefore, the trapping bias may result from increased trap encounter due to more rapid movement by A. americanum, although greater attraction to carbon dioxide may also play a role. Tick traps are useful for intraspecific between-habitat comparisons. Early in their seasonal activity period, larval I. dammini were better represented in collections from mouse hosts than in flagging samples. Apparently, sampling from favored hosts can detect ticks at low population levels, but often cannot be used to get accurate estimates of pathogen prevalence in questing ticks.