Myofibrillar proteins assemble to form the highly ordered repetitive contractile structural unit known as a sarcomere. Studies of myogenesis in vertebrate cell culture and embryonic developmental systems have identified some of the processes involved during sarcomere formation. However, isoform changes during vertebrate muscle development and a lack of mutants have made it difficult to determine how these proteins assemble to form sarcomeres. The indirect flight muscles (IFMs) of Drosophila provide a unique genetic system with which to study myofibrillogenesis in vivo. We show in this paper that neither sarcomeric myosin nor actin are required for myoblast fusion or the subsequent morphogenesis of muscle fibres, i.e. fibre morphogenesis does not depend on myofibrillogenesis. However, fibre formation and myofibrillogenesis are very sensitive to the interactions between the sarcomeric proteins. A troponin I (TnI) mutation, hdp³, leads to an absence of TnI in the IFMs and tergal depressor of trochanter (TDT) muscles due to a transcript-splicing defect. Sarcomeres do not form and the muscles degenerate. TnI is part of the thin filament troponin complex which regulates muscle contraction. The effects of the hdp³ mutation are probably caused by unregulated acto-myosin interactions between the thin and thick filaments as they assemble. We have tested this proposal by using a transgenic myosin construct to remove the force-producing myosin heads. The defects in sarcomeric organisation and fibre degeneration in hdp³ IFMs are suppressed, although not completely, indicating the need for inhibition of muscle contraction during muscle development. We show that mRNA and translated protein products of all the major thin filament proteins are reduced in hdp³ muscles and discuss how this and previous studies of thin filament protein mutants indicate a common co-ordinated control mechanism that may be the primary cause of the muscle defects.