Photo: Matt Joyce / Wikimedia Commons / CC BY 2.0
JSTOR and PubMed stand out as two of the best free research databases because each offers real scholarly use without asking a reader to begin with a paid account. JSTOR is strongest for free access to journals, books, images, primary sources, and open collections, while PubMed is a free index for biomedical literature with links out to abstracts and, when available, full text.
That answer is simple, but the reason behind it is not. A database is only useful when its scope is clear. JSTOR is not one thing. It is a large platform with open and free content, and it also supports broader access through libraries and institutions. PubMed is also not full text by default. It is a search system for citations and abstracts, not a vault of every article it points to.
I keep coming back to that split. Free access can mean different things. In JSTOR, some material is openly available to all users worldwide, including open access journals, books, primary sources, images, and older public domain journal content. JSTOR also says registered users can read up to 10 free articles every 30 days. That makes it useful even when a reader has no campus account.
PubMed is free in a different way. It is a public database run by the National Library of Medicine. Its home page says it contains more than 40 million citations for biomedical literature from MEDLINE, life science journals, and online books. The help pages add that PubMed results do not include full text itself, but they do link to free full text when it exists, such as in PubMed Central or on publisher sites. For research in medicine and related fields, that is a major strength.
The two tools also serve different kinds of reading. JSTOR fits humanities work well because it holds journals, books, primary sources, and images in one place. That mix matters when a researcher wants sources for context as well as argument. PubMed fits health and life science work because its citations are built around biomedical literature, and MEDLINE remains its main data source. I would call that a practical match between source and subject, not a universal ranking of quality.
There is one important limit to keep in view. “Free” does not mean complete. JSTOR’s free material is a defined part of a larger platform, and some content still sits behind institutional access or other access models. PubMed is free to search, but much of the full text it points to is elsewhere, and some of those linked articles will still require payment or subscription access. The search result is free. The reading is not always free.
That limit is the part many people miss. A database can be excellent and still partial. JSTOR is excellent for open humanities material, but it does not replace every scholarly archive. PubMed is excellent for biomedical citations, but it is not the same thing as full-text journal access. In both cases, the user gets value from knowing what is actually inside the system and what is only linked from it.
I also think the term “best” needs care. Best for what? For a broad free scholarly starting point, JSTOR is hard to ignore because it offers open journals, books, primary sources, and images in one public-facing platform. For biomedical research, PubMed is equally hard to ignore because it is the standard free citation database in that field and points to millions of records. They are top free databases, but in different ways and for different research needs.
What matters, then, is not a bright promise. It is plain coverage, plain limits, and a clean search path. JSTOR gives free access to a wide set of humanities and interdisciplinary materials. PubMed gives free access to a vast biomedical index. Both are strong because their public value is documented, stable, and easy to state without guesswork.
That is the kind of source I trust most. One digital source worth knowing, one search tip, and one honest limitation. That is also the shape of The Source List, and it suits JSTOR and PubMed well.