The structure of the gene encoding the apoprotein of phytochrome B (PHYB1) in tomato has been determined from genomic and cDNA sequences. In contrast to PHYA, PHYB1 lacks an intron upstream of the first ATG. A single transcription start site was found by 5' RACE at -116. Tomato PHYB1 spans 7 kb starting from the first ATG. The coding region is organized into four exons as for other angiosperm PHY. The deduced apoprotein consists of 1131 amino acids, with a molecular mass of 125.4 kDa. Tomato phytochrome B1 shares 78% and 74% identity with Arabidopsis phytochromes B and D, respectively. Along with the normally spliced full-length transcripts, sequences of reverse transcriptase-PCR clones revealed five types of alternative transcripts. Each type of alternative transcript was missing a considerable part of the coding region, including the chromophore-binding site. The four putative PHYB1 mutants in tomato, which are temporarily red-light insensitive (tri), were each confirmed to have a mutation in PHYB1. Each mutation arose from a different, single-base substitution. Allele tri1 is presumably a null because the mutation introduces a stop at codon 92. In tri3, val-238 is replaced by Phe. The importance of this valine residue is evidenced by the fact that the tri3 phenotype is as strong as that of tri1. Alleles tri2 and tri4 encode proteins truncated at their C-termini. The former lacks either 170 or 438 amino acids, depending upon which of two types of splicing occurs during transcript maturation, while the latter lacks 225.