This is a report of many distant but significant protein sequence relationships between human proteins and transposable elements (TEs). The libraries of human repeated sequences contain the DNA sequences of many TEs. These were translated in all reading frames, ignoring stop codons, and were used as amino acid sequence probes to search with BLASTP for similar sequences in a library of 25,193 human proteins. The probes show regions of significant amino acid sequence similarity to 1,950 different human genes, with an expectation of <10 ؊3 . In comparison with previous REPEATMASKER (Institute for Systems Biology, Seattle) studies, these probes detect many more TE sequences in more human coding sequences with greater length than previous work using DNA sequences. If the criterion is opened, very many matches are found occurring on 4,653 different genes after correction for the number seen with random amino acid sequence probes. The processes that led to these extensive sets of sequence relationships between TEs and coding sequences of human genes have been a major source of variation and novel genes during evolution. This paper lists the number of sequence similarities seen by amino acid sequence comparison, which is surely an underestimate of the actual number of significant relationships. It appears that many of these are the result of past events of duplication of genes or gene regions, rather than a direct result of TE insertion. This report of observable relationships leaves to the future the functional implications as well as the detection of the events of TE insertion.sequence similarity ͉ mobile elements ͉ genes ͉ repeated sequences T he contribution of transposable elements (TEs) to eukaryotic genomes has been much studied and will not be reviewed here. I use the phrase TE to refer to any of the known repeated sequences, not including short tandem repeats and simple sequences. The presumption is that most TEs are the product of past insertions of active TE family members into the genomes of many eukaryote ancestors. No uncertainty in meaning results for the minority for which this may not have been demonstrated. Usage does not imply that the elements are currently active and transposable.This article describes measurements of the relationships of TE sequences to human coding sequences, which probably represent much of the contribution TEs have made to the coding sequences of human proteins. The logic underlying this approach is that among the human ''repeated sequences,'' there are many sequences that are related to TEs but are mostly incomplete and damaged, and the original ORFs of the once active TE cannot be recognized in many cases. Therefore, I have simply taken as probes the translation in each of the possible reading frames, ignoring stop codons, to make a set of six probes for each of the repeated sequences. The sequences available may well have undergone various mutations, including frame-shift events, and these will have broken up and limited the length of the TE probes. The relationship of the pr...