Several techniques were trialed for monitoring the newly-established fur seal haul-out site at Steamers Head, NSW. In order to minimise our impact, we also monitored for any disturbance to seals from our sampling. The site is unusual as while it can only be observed by boat seals perch on a steep collection of ledges, which means that regardless of their thigmotatic nature, counts can be made of seals from a distance on the horizontal plain. Counts were performed using three methods and at four distances. The three methods trialed were direct counts through binoculars, counts from video footage and counts from projected photographs. Originally, four distances from the haul-out were determined with a laser rangefinder as 150 metres, 100 metres, 75 metres and 50 metres. Disturbance was ranked and recorded, which led us to quickly abandon our 50 metre observation distance. Both binoculars and video techniques were reliable methods for counts of all seals, however; only the binocular technique was reliable for identification of different species and age groups. Counts from photographs were unreliable. Counts through binocular are recommended for long term monitoring as they provide, overall, the most accurate counts of different age classes and a precise repeatable measure. Also, unlike the other methods, binocular counts do not require post-field processing. Monitoring of our disturbance to the seals tentatively found that hauled seals were more easily disturbed when the boat was closer to the haul-out. Disturbance occurred at greater distances as the haul-out seasonally expanded in numbers. We recommend, for this haul-out, that to perform counts unbiased by the impact of observations, researchers must remain at least 75 metres from the haul-out when there are fewer than 50 seals, and at least 100 metres when there are 50 or more seals.