Two glasshouse experiments were conducted to evaluate the genotypic variation amongst cereal genotypes in phosphorus uptake from relatively insoluble iron phosphate. Optimum rates of iron phosphate were established by growing 3 wheat and 1 triticale genotype on an infertile sand amended with iron phosphate. Shoot dry weight of all genotypes showed a classic Mitcherlich response with 95% maximum growth achieved with 174 mg P/kg soil. Two rates of FePO4 were selected representing a deficient and sufficient supply (26 and 339 mg P/kg soil, respectively).These rates were used to screen 99 wheat, 8 triticale, and 4 cereal rye genotypes for phosphorus-use efficiency. Phosphorus efficiency was rated by 4 criteria: shoot dry weight at deficient P supply, shoot weight at deficient supply relative to shoot weight at sufficient P supply, P uptake efficiency (amount of P taken up per unit of P supplied), and P utilisation efficiency (shoot weight per unit P in plant).No genotypes were rated as efficient under all 4 criteria. Only 2 genotypes were rated efficient (rye Bevy, rye PC00361) and one inefficient (Machete) under 3 criteria. Seven genotypes were rated as efficient on 2 indices (wheat Chinese 80-55, Westonia, and Wawht 2147; triticale Treat, AT48-94, and TX93-78-1; rye Bulgarian Pento), whereas 7 genotypes were rated as inefficient on 2 indices (Boricuta, Cadoux, Cunderdin, Insignia, Kalingri, Perenjori, and triticale Abacus). Significant genotypic variation was identified in cereals in the ability to take up and utilise P from poorly soluble Fe-P, although all genotypes were able to utilise this source of phosphorus to some degree.