Less than 4,000 yellow-billed loons (Gavia adamsii) breed in remote and disjunct locations in northern Alaska, USA. Over 75% of the United States population of yellow-billed loons nests in the National Petroleum Reserve-Alaska (NPRA), where impending oil and gas development will intersect their breeding range. We investigated the relationship of recent oilfield development to occupancy of yellow-billed loon territories by breeding pairs (indicated by active nests) and broods using 14 years of aerial surveys on the Colville River delta. We also evaluated the survey requirements prescribed by the Bureau of Land Management (BLM) for NPRA. We began aerial surveys for yellow-billed loons in 1993, prior to construction of the Alpine oilfield in 1998, and followed territories through 2008, after construction of 2 additional satellite drill sites. We used records from 37 breeding territories on 36 lakes in model selection analyses to examine how habitat and disturbance factors (proximity to facilities and construction time period) influenced occupancy by breeding pairs and broods. Annually, 13 AE 2.5 (SE)% (n ¼ 14 yr) of broods (n ¼ 19) moved from nesting lakes to adjacent brood-rearing lakes, and the remainder stayed in nesting lakes (n ¼ 128). Lakes used for nesting and brood-rearing were almost 25 times larger ( x ¼ 95.9 AE 25 ha, n ¼ 23 lakes) than nesting lakes from which broods left ( x ¼ 4.0 AE 1.1 ha, n ¼ 7 lakes, P < 0.001). Thirty-eight percent of territories (n ¼ 14 territories) were on lakes shared by >1 breeding pair. Lake type (deep open lakes with islands or polygonized margins, deep open lakes without islands or polygonized margins, and tapped lakes with high-water connections) was the most influential covariate on occupancy by breeding pairs, and lake area was most influential on occupancy by broods. Time period and distance to facilities (as discrete zones at 1.6 km and 3.2 km and as linear distance) were factors in the highest-ranked models for 5 of the 6 model sets that included disturbance parameters. Interaction terms for time period and distance to oilfield facilities were factors in 3 of 6 model sets. The pattern of occupancy of breeding territories, however, was not consistent with disturbance-related effects. Occupancy of territories by breeding pairs was lower in the predevelopment period (lowest human activity) than in the latest development period (highest human activity) and higher in the zones near oilfield facilities than far from facilities. Occupancy of territories by broods was highest in the latest development period and similarly high in zones near and far from facilities. Application of BLM minimum survey requirements (3 yr with 2 surveys/yr) to the initial 3 years of surveys in this study resulted in detecting 81% of the known territories on the Colville River delta. The BLM restrictions on development were judged conservative in maintaining breeding territories around oilfield developments. Our results did not demonstrate displacement of nests or broods from long-standing territori...