For a complementary view from the protein side of things, see proteinatlas.org. This is a beautiful site that organizes information about where proteins are expressed, as quantified by antibody staining. For an example, see their entry on Hras. You can drill down to see the actual stains for any given sample, e.g. this stain of Hras in the colon
In addition to the resources suggested here, you may also take a look at "Expression" tab in HPRD / Human Proteinpedia. The reason to check in HPRD is the data is manually curated from literature and you can get a reference.
You can have a look at the Gene Expression Atlas hosted b EMBL-EBI. [?]You can enter you gene(s) of interest, the organism and the conditions of the experiments.[?] You should be able to customize the results if you try to use the Gene Expression Atlas REST APIs. It can return you a result either in XML or JSON format.
I would also recommend:
ImmGen (human and mouse)
Finally, you can also use GEO Profiles using operators in the search field (with the preview/Index tab):
(BCL2[Gene Symbol]) AND (30[Min Value Rank]) ...
I'll provide an answer that I have given in the past - BLAST search against an EST database. There are more tissues that served as sources for the cDNA libraries than you'll find in many if not all of the databases listed above. (These are excellent choices for one approach - I'm offering another). Look to see from which tissue the library was made, from which the clones were sequenced to give the BLAST hit(s) you observe.
At the same time, you may want to consider RNAseq data. There is not so much readily available compared to other data types and databases, but you should think about ways to identify expression of a given gene by looking at this ever increasing data type.