1. Increased myoplasmic inorganic phosphate (P(i)) has been suggested to have an important role in skeletal muscle fatigue, especially in the early phase. In the present study we used intact fast-twitch muscle cells from mice completely deficient in creatine kinase (CK(-/-)) to test this suggestion. These CK(-/-) muscle cells provide a good model since they display a higher P(i) concentration in the unfatigued state and fatigue without significant increase of P(i). 2. Tetanic contractions (350 ms duration) were produced in intact single muscle fibres. The free myoplasmic [Ca(2+)] ([Ca(2+)](i)) was measured with the fluorescent indicator indo-1. The force-[Ca(2+)](i) relationship was constructed from tetani at different frequencies. 3. Compared with wild-type fibres, CK(-/-) fibres displayed lower force in 100 Hz tetani and at saturating [Ca(2+)](i) (i.e. 100 Hz stimulation during caffeine exposure), higher tetanic [Ca(2+)](i) during the first 100 ms of tetanic stimulation, reduced myofibrillar Ca(2+) sensitivity when measurements were performed 100-200 ms into tetani, and slowed force relaxation that was due to altered cross-bridge kinetics rather than delayed Ca(2+) removal from the myoplasm. 4. In wild-type fibres, a series of 10 tetani resulted in reduced tetanic force, slowed force relaxation, and increased amplitude of [Ca(2+)](i) tails after tetani. None of these changes were observed in CK(-/-) fibres. 5. Complementary experiments on isolated fast-twitch extensor digitorum longus muscles showed a reduction of tetanic force and relaxation speed in CK(-/-) muscles similar to those observed in single fibres. 6. In conclusion, increased P(i) concentration can explain changes observed in the early phase of skeletal muscle fatigue. Increased P(i) appears to be involved in both fatigue-induced changes of cross-bridge function and SR Ca(2+) handling.