Are you interested in the prediction of promoter regions or just the prediction of cis-regulatory elements? For the former, you can check this compilation of resources: http://www.shodhaka.com/cgi-bin/startbioinfo/simpleresources.pl?tn=Promoter%20prediction
Of course this prediction has its limitations, since not all promoters in eucaryotes have the same characteristic elements (for example, not all of them have TATA box), but some of the limitations are mentioned in the compilation.
If looking for cis-regulatory elements in a promoter region, you can use the databases mentioned by Daniel, and also give a try to Transfac http://www.gene-regulation.com/pub/databases.html or PLACE, if working with plant sequences http://www.dna.affrc.go.jp/PLACE/
However, all these databases just look for regulatory elements that have already been reported in a particular promoter in response to a particular stimulus. However, transcription factors can sometimes be promiscuous and are able to bind slightly different sequences, or, when the element is in combination with another element, sequence restrictions can change. In these cases maybe the search for already known elements won't be very useful.
For de novo prediction of regulatory elements, I recommend using (if available) sequences of orthologous promoters or promoters of genes co-expressed with your gene of interest. In those cases you would expect they share functional regulatory elements that are evolutionarily conserved. Having those sequences you can use SCOPE http://genie.dartmouth.edu/scope/ or Melina http://www.openhelix.com/cgi/tutorialInfo.cgi?id=26
Try the mammalian promoter database, http://mpromdb.wistar.upenn.edu/
the eukaryotic promoter database, http://epd.vital-it.ch/
or the Transcriptional Regulatory Element Database, http://rulai.cshl.edu/cgi-bin/TRED/tred.cgi?process=home
If you are working with a non-mammalian species, the resources would change, so please make mention of it.
The major limitation of these databases is that in addition to proximal promoters, enhancers can also play a major role in gene regulation, and it is not trivial to map them to your gene of interest. Nonetheless, these databases can be a good starting point for reporter assays or for looking at potential motifs/TFs that might regulate a target gene of interest.
Are you interested in the prediction of promoter regions or just the prediction of cis-regulatory elements? For the former, you can check this compilation of resources: http://www.shodhaka.com/cgi-bin/startbioinfo/simpleresources.pl?tn=Promoter%20prediction
Of course this prediction has its limitations, since not all promoters in eucaryotes have the same characteristic elements (for example, not all of them have TATA box), but some of the limitations are mentioned in the compilation.
If looking for cis-regulatory elements in a promoter region, you can use the databases mentioned by Daniel, and also give a try to Transfac http://www.gene-regulation.com/pub/databases.html or PLACE, if working with plant sequences http://www.dna.affrc.go.jp/PLACE/
However, all these databases just look for regulatory elements that have already been reported in a particular promoter in response to a particular stimulus. However, transcription factors can sometimes be promiscuous and are able to bind slightly different sequences, or, when the element is in combination with another element, sequence restrictions can change. In these cases maybe the search for already known elements won't be very useful.
For de novo prediction of regulatory elements, I recommend using (if available) sequences of orthologous promoters or promoters of genes co-expressed with your gene of interest. In those cases you would expect they share functional regulatory elements that are evolutionarily conserved. Having those sequences you can use SCOPE http://genie.dartmouth.edu/scope/ or Melina http://www.openhelix.com/cgi/tutorialInfo.cgi?id=26
try www.genomatix.de/matinspector ; this software is free. More complex softwares provided by genomatix allow the comparison between species, for example "eldorado" but you should buy them