Methodological aspects of determination of P availability in ducks were addressed in this study. Three balance experiments with quantitative excreta collection were performed with growing male White Pekin ducks that were 3 wk of age. When P from monobasic calcium phosphate (MCP) was added in graded levels below the requirement to a low-P basal diet, ducks responded nonlinearly in P accretion. With increases in dietary Ca:P ratio from 1.2:1 to 1.6:1 to 2.0:1, y(max) in P accretion increased from 0.91 to 0.97 to 1.03 g/duck during the 5-d period of measurement, and marginal efficiency of utilization (Deltay/Deltax) of P from MCP increased from 75 to 85 to 91%. The overall utilization of plant P from the low-P basal diet was adversely affected by increasing dietary Ca:P ratio and decreased from 49 to 45 to 40%. Seven inorganic P sources were supplemented to a low-P basal diet to increase P content by 0.9 or 1.9 g/kg. Utilization of P from the sources was lower at 1.9 g/kg compared with at 0.9 g/kg, although supply was still below requirement, indicating that ducks fed 1.9 g/ kg were in the diminishing part of the response curve. Thus, only the values determined at the low level of supplementation were considered as a measure of availability. P availabilities were 100% for monosodium phosphate, 96% for anhydrous dibasic calcium phosphate, 92% for calcium sodium magnesium phosphate, 91% for monodibasic calcium phosphate, 86% for 2 batches of dihydrated dibasic calcium phosphate, and 77% for calcium sodium phosphate. It was concluded that availability studies aiming to characterize the capacity of an ingredient require standardization with regard to the level of P supply and the dietary Ca:P ratio. Similar to results from other species, the availability of inorganic feed phosphates for ducks is on an overall high level, but differences between P sources should be considered in diet formulation.