A common method for sampling tick populations is flagging, which is a method of dragging a white cloth over a plant substrate for a fixed distance along a transect. Flagging over rough physical surfaces or using long subtransect lengths could lead to the underestimation of tick densities. Using estimates of the drop-off rates of adult Ixodes pacificus in flag sampling, optimal sampling schemes (the length and number of subtransects) were examined using the relationships between the tick drop-off rates (c), the tick density and distribution patterns and the roughness of the sampled plant substrate. It was found that the optimal number of subtransects and Lmax, the longest subtransect length which did not significantly underestimate the tick density from c, were affected by the tick density, substrate roughness and tick distribution pattern. This study also showed that the density and distribution of I. pacificus and Dermacentor occidentalis varied greatly over time in the populations sampled, while the Dermacentor variabilis densities were low and showed no significant changes over time. Both I. pacificus and D. occidentalis had clumped distributions along trails and these clumps were aggregated. However, the clump sizes (or individual clump areas) changed significantly over time because of density fluctuation or the movement of ticks. Finally, a positive association between the number of I. pacificus and D. occidentalis adults was observed from 2 m subtransect flag collections in March 1995; no relationship was found in 1994.