ABSTRACT
From a systematic study of 1009 references in The Astrophysical Journal it was found that 12.2% had errors. Only 0.4% of the referenced papers could not be found at all; another 3.0% were found by searching in volume, annual, and five-year indices and the remaining correctly and incorrectly referenced papers were found right away. Another 8.3% of the references have errors in the first authors' names or in the journal names, and volume and page numbers such that they could be misplaced in the Science Citation Index (S CI). However, the compilers of SCI match all citations against a computerized file of the source papers and correct some of the citations, so that only 3.6% of the citations are missing or displaced in SCI.