Phosphorus fertiliser is necessary to maintain the production of New Zealand hill country pastures. Technological advances in GIS and GPS have given aerial fertiliser applicators more precision than ever before. The key question is; to what extent do soil chemical properties vary within hill country landscapes, and is a knowledge of the variability able to be used to differentially apply fertiliser, maximise P fertiliser use efficiency, and minimise P loss to waterways? The objective of this thesis was to quantify the phosphorus variability on Blue Duck Station, a summer-moist hill-country farm located in the Central North Island of New Zealand. The soil comprised Allophanic Soil (Andisols) on the flat to low slopes with Brown Soil (Inceptisols) on the steeper hills. Blue Duck Station was delineated into six soil-landscape groups (north facing medium slopes, north facing steep slopes, south facing steep slopes, south facing medium slopes, flat ridge tops, and flat valley floors, based on slope, aspect, and elevation using GIS. Three replicate units were randomly selected from each of the six soil-landscape groups, which gave 18 study-sites in total. For each study site, one transect sample and five grid samples were collected. Olsen P, Anion Storage Capacity, pH, Sulphate-sulphur and Cations (Mg, Na, K, Ca) were determined. The Olsen P ranged from a mean of 6.8 in the south facing medium slopes to 21.2 in the north facing medium slopes. North facing medium slopes and valley low floors had higher mean Olsen P values than all other landscape groups (P<0.01) and also had the greatest variability in Olsen P values. The majority of Olsen P values were considered ‘low’ or ‘very low’. Soil pH was consistent, ranging between 5 and 5.4, whilst anion storage capacity (ASC) levels ranged from 35% to 85%. High ASC levels indicated soils of volcanic origin, containing tephra, whereas low ASC levels were associated with soil predominantly derived from tertiary sedimentary rock. At Blue Duck Station, the application of higher rates of P fertiliser to south facing steep slopes and south facing medium slopes, with medium rates of P fertiliser addition to north facing steep slopes and ridge low slopes would optimise fertiliser use. There is limited benefit in applying fertiliser to north medium slopes and flat valley floors as Olsen P is near optimal for this environment and stock transfer will continue to add nutrients to these areas. The variability within a sample site (between grid and transect samples) was often greater than the range of variability the means across all the landscape groups. Due to such large sample variability, difficulties arise when attempting to recommend accurate fertiliser regime, or sampling programs. The soil analysis results from a small number of samples should be treated with caution. The Best Management Practice for soil sampling the Central North Island Hill Country would be to avoid sampling small exceptional areas like flat valley floors and flat ridge tops. Focusing sample regimes on a number of transects across northern and southern slopes would give more representative results.