The sequencing of the zebrafish genome should be completed by the end of 2002. Direct assignment of function on the basis of this information would be facilitated by the development of a rapid, targeted 'knockdown' technology in this model vertebrate. We show here that antisense, morpholino-modified oligonucleotides (morpholinos) are effective and specific translational inhibitors in zebrafish. We generated phenocopies of mutations of the genes no tail (ref. 2), chordin (ref. 3), one-eyed-pinhead (ref. 4), nacre (ref. 5) and sparse (ref. 6), removing gene function from maternal through post-segmentation and organogenesis developmental stages. We blocked expression from a ubiquitous green fluorescent protein (GFP) transgene, showing that, unlike tissue-restricted limitations found with RNA-based interference in the nematode, all zebrafish cells readily respond to this technique. We also developed also morpholino-based zebrafish models of human disease. Morpholinos targeted to the uroporphyrinogen decarboxylase gene result in embryos with hepatoerythropoietic porphyria. We also used morpholinos for the determination of new gene functions. We showed that embryos with reduced sonic hedgehog (ref. 9) signalling and reduced tiggy-winkle hedgehog (ref. 10) function exhibit partial cyclopia and other specific midline abnormalities, providing a zebrafish genetic model for the common human disorder holoprosencephaly. Conserved vertebrate processes and diseases are now amenable to a systematic, in vivo, reverse-genetic paradigm using zebrafish embryos.