Disturbances to harbor seals, Phoca vitulina richardsi, during 1991 and 1992 pupping seasons were observed at Puffin Island, Clements Reef, and Skipjack Island in Washington state. Harassment (≥ one seal entering the water) of seals ashore was common (≥71% of survey days) and primarily caused by powerboat operators approaching to observe seals. Recovery (number of seals on a haul-out site returned to preharassment levels) following a harassment was less at Puffin Island (19%) than at Clements Reef (54%) and Skipjack Island (45%). Additionally, seals were more vigilant (P<0.003) at Puffin Island than at the other two locations. These results indicated that seals at Puffin Island were less tolerant of disturbance than seals at other sites. This could possibly be attributed to a greater (P<0.05) percentage of pups ashore (17%) than at Clements Reef (3%) and Skipjack Island (3%). Because of this, we expected that powerboats would disturb seals from greater distances at Puffin Island. To test this, we used a theodolite to determine distance between seals and an approaching vessel at Puffin Island and Clements Reef. There was, however, no significant (P>0.05) difference in distances at which disturbances occurred. The most notable difference in distance of disturbance was between initial and subsequent harassments during a haul-out period. Those seals remaining or returning to shore after a harassment were more tolerant of powerboats, allowing significantly (P<0.05) closer approaches than those initially harassed. Seals detected (head raised and oriented toward the potential disturbance) a powerboat at a mean distance of 264 m, and harassments occurred when boats approached, on average, to within 144 m. Results of this study exemplify the variability in reaction to disturbance and the necessity for considering these differences for minimizing disturbance.