The structural organization of the chromosomal gene for human parvalbumin was determined mostly by sequencing exons and intron exon junctions of a 7500 base-pair (bp) long genomic clone derived from a chromosome 22-specific gene library. Four exons coding for 100 from a total of 109 amino acids were detected in this clone and 472 bp of the 5'-flanking region were sequenced. The region corresponding to the C-terminal amino acids 101 to 109 of human parvalbumin was determined by sequencing a cDNA fragment derived from human brain mRNA after amplification by the polymerase chain reaction. The first intron is placed 7 bp upstream from the ATG translation start signal, whereas all other splice sites divide putative Ca2+-binding domains. All intron positions coincide exactly with those reported for the rat parvalbumin gene. The 5' mRNA leader sequence has a similarity of 57%, the coding region of 91% and the 3' non-coding region of 83% to the corresponding rat sequences. Only nine conservative amino acid replacements were observed between human and rat parvalbumins. The predicted secondary structures for human, rat, mouse and rabbit parvalbumins are very similar, indicating a strong structural relationship among mammalian parvalbumins. Several elements with potential transcription regulatory activities were found in the region immediately 5' to the transcription start site including a TATA box (TATATA) and a CAAT box (CCAAAAT). Several regions in the putative promoter are strongly conserved between the human and rat parvalbumin genes. One of these with a length of 32 bp is identical with the rat counterpart and has a high degree of homology to a promoter region in the myosin light chain 3F gene, which is expressed in fast contracting/relaxing muscle fibers (anaerobic/type IIb), the cell type that also exhibits highest levels of parvalbumin expression. The human parvalbumin mRNA contains the putative polyadenylation signal AATAAA 13 nucleotides upstream from the polyadenylation site. A 700-nucleotide long parvalbumin mRNA is synthesized at low levels in the human cerebellum as well as in the neuroblastoma cell line SK-N-BE.