Using degenerate primers based on published beta 2-microglobulin sequences we were able to obtain an expected 111 base pairs (bp) polymerase chain reaction (PCR) fragment from tilapia genomic DNA. The sequence of this fragment showed a high degree of similarity to mouse beta 2-microglobulin at the protein level. We used these primers in an "anchored PCR" to obtain a 213 bp PCR fragment from a carp cDNA library. This was then used to clone a full-length beta 2-microglobulin cDNA from carp. The carp sequence showed the highest similarity to rabbit beta 2-microglobulin. Both sequences showed strong similarities to all previously published vertebrate beta 2-microglobulin sequences. The predicted protein secondary structure of both the carp and tilapia clones was almost identical to the corresponding regions of previously known vertebrate beta 2-microglobulin protein sequences. When either the carp or tilapia probes were used against corresponding northern blots, they hybridized to a message of approximately 800-1000 bases long, which corresponds to the previously published lengths of beta 2-microglobulin mRNAs. Southern blotting indicated that beta 2-microglobulin was encoded by a single copy gene in both cases. Phylogenetic analysis indicated that the sequences were related to the beta 2-microglobulins of higher vertebrates but grouped together in an ancestral position.