Data on nymphal Ixodes scapularis ticks submitted by the public to the University of Rhode Island Tick Research Laboratory for testing from 1991 to 2000 were compared with human case data from the Rhode Island Department of Health to determine the efficacy of passive tick surveillance at assessing human risk of Lyme disease. Numbers of ticks submitted were highly correlated with human cases by county (r = 0.998, n = 5 counties) and by town (r = 0.916, n = 37 towns), as were the numbers of positive ticks submitted (r = 0.989 by county, r = 0.787 by town). Human cases were correlated with ticks submitted by town each year, and with positive ticks in all but 2 years. Thus, passive tick surveillance effectively assessed geographical risk of human Lyme disease. In contrast, tick submissions through time were not correlated with human cases from year to year. Dog seropositivity was significantly correlated with human cases by county in both years tested, but by town in only one of two years. Numbers of ticks submitted were correlated with dog seropositivity by county but not by town, apparently because of high variability among towns with small sample sizes. Our results suggest that passive tick surveillance, using ticks submitted by the public for Lyme spirochete testing, can be used to assess the geographical distribution of Lyme disease risk, but cannot reliably predict Lyme incidence from year to year.