Information on potassium forms in calcareous, partly saline date palm orchard soils of Bushehr Province, and on how common extractants compare with ammonium acetate there, is limited. This study determined potassium forms and compared potassium extracted by seven extractants against 1 M ammonium acetate. Thirty soil samples from 15 orchards in Bushehr, Dashtestan, and Tangestan counties (0–30 and 30–60 cm) were analyzed for soluble, exchangeable, non-exchangeable, and structural potassium. The soils were calcareous (calcium carbonate equivalent 60–90%), mainly sandy loam, silt loam and clay loam in the surface layer and loam, silt loam, clay loam and sand in the subsurface layer, low in organic matter and cation exchange capacity, and often moderately to highly saline. Soluble, exchangeable, non-exchangeable, structural, and total potassium ranged 13.9–30.8, 58.1–326.8, 97.3–714, 1308–3334, and 2246–4184 mg kg⁻¹, respectively. Structural potassium (about 77%) was the largest pool, followed by non-exchangeable (16%), exchangeable (6%), and soluble (1%). Non-exchangeable potassium reached about 700 mg kg⁻¹ in some light- to medium-textured soils; its mineral origin remains unknown without mineralogical data. Morgan extracted the most potassium and correlated very highly with 1 M ammonium acetate (r = 0.984 and 0.982 in surface and subsurface layers) and AB-DTPA; 0.01 M calcium chloride extracted the least. As extractants were compared with ammonium acetate, not plant response, correlations show agreement with the reference only. Morgan is a candidate alternative with practical advantages over ammonium acetate (shorter equilibration time, multi-element use), but its suitability requires calibration against plant uptake.