The tule elk (Cervus elaphus nannodes) was reintroduced to Point Reyes National Seashore, California in 1978. After exhibiting irruptive growth for 20 years, elk abundances have fluctuated. Three subpopulations of elk currently inhabit Pt. Reyes: a fenced subpopulation at high density; a free-ranging subpopulation in designated wilderness at low density; and a freeranging subpopulation on active ranchland at low density. Little is known about the spatial ecology and population dynamics of tule elk. This unusual situation provided an opportunity to quantify and compare factors affecting the home ranges, group sizes, resource selection, smallscale movements and population dynamics of tule elk. I collected biweekly fixes on radiocollared elk cows between 2005 and 2008 to examine herd structures, space use patterns, grouping behavior and survival rates of tule elk. Radio-telemetry data revealed 6 herds of elk at Pt. Reyes (4 at Tomales), which had distinct herd ranges and showed little spatial interaction. Fixed kernel home ranges were some of the smallest for elk in North America (217 ha, SE = 35), and were negatively correlated with elk density. Home range sizes were 300% smaller than previous estimated for elk at Pt. Reyes in 1998. The sizes of elk groups peaked during the winter and mid-summer, and were smallest during the spring parturition and fall rutting periods. Elk with larger home range sizes generally formed smaller groups. Resource selection function (RSF) results indicated that tule elk at Pt. Reyes selected flat, grassland-dominated habitats over hilly, scrub and forest habitats; although resource selection patterns varied among herds. Elk shifted to lower elevations, steeper slopes and more scrub-dominated habitats during the dry season (May – October). Most elk showed an aversion to areas close to roads and trails, but this result was not consistent across all herds. I quantified small-scale movement patterns of tule elk using hourly fixes from GPS-collared elk. Elk exhibited daily cycles in their movement patterns by increased their hourly movements in the morning and evening hours. Elk daily movements were longest during wet spring months and shortest during the dry fall months. Elk slowed when traveling through homogeneous grassland habitat. Windy conditions caused elk to slow, but precipitation had no effect on elk movement. Elk moved faster when closer to roads and trails. Elk in higher density herds moved slower on an hourly scale than lower density herds, but herd density had no affect on daily movement patterns. Using known-fate models applied to radiotelemetry data, I estimated annual cow elk survival was 0.96 (SE = 0.02). I captured and monitored radio-collared elk calves from 2005 to 2007 to estimate calf survival and causes of mortality. Using known-fate models, I estimated annual survival for elk calves was 0.81 (SE = 0.02). The primary cause of death for both cows and calves was starvation, which was often accompanied by copper and selenium deficiencies. Sensitivity and elasticity estimates of elk vital rates revealed that adult cow elk survival contributed the most (30%) to elk annual population change (λ). The D Ranch and Limantour herds were expected irruptive growth and by 2018 increase to 393 and 389 elk, respectively, based on stochastic population projection models. The projected future abundance of elk in the Tomales herds was dependant on the previous year’s annual rainfall: models projected increases in elk abundance following years of above-average rainfall but no change in elk abundance following years of below-average rainfall. Future growth of the D Ranch and Limantour elk herds will likely concentrate on adjacent ranchlands, given RSF models results, and will likely result in future conflicts between NPS management, unless proactive actions are taken.