How to Write a Literature Review (Step-by-Step)

A literature review is not a summary of everything published on a topic. That is a bibliography. A literature review is a critical argument: it explains what has been studied, how different approaches have contributed to understanding, where the debates and disagreements are, and what remains unresolved or unexplored. Done well, a literature review justifies your research question by showing precisely where it fits and why it matters.

This guide explains how to write a literature review from scratch: how to search for sources, how to organize what you find, how to synthesize rather than summarize, and how to write the review itself.

What a Literature Review Is (and Is Not)

Before writing anything, it helps to be clear about what you are actually producing.

A literature review is:

• A critical synthesis of existing scholarship on a topic
• An argument about the state of the field
• A demonstration that you understand what has been studied and how
• A setup for the gap your research fills

A literature review is not:

• A list of summaries: "Smith (2015) says X. Jones (2018) says Y."
• A chronological history of all publications on a topic
• An exhaustive catalog of every paper ever written on a subject
• A neutral presentation with no analytical position

The most common mistake in literature review writing is producing an annotated bibliography: a series of individual summaries that never speak to each other. A literature review should read as an argument, not as a list.

Step 1: Define Your Scope

Before you search for sources, define the boundaries of your review.

Ask:

• Topic scope: What aspect of the broader subject are you reviewing? "Anxiety" is too broad. "Anxiety in doctoral students and its relationship to advisor supervision quality" is scoped appropriately.
• Time scope: Are you reviewing all historical literature or recent work? For most academic literature reviews, the last 10 to 20 years is the primary range, with foundational older work cited for context.
• Disciplinary scope: Are you reviewing only your home discipline or relevant work from adjacent fields? Interdisciplinary review adds richness but also scope creep.
• Geographic or contextual scope: Is your review limited to a specific context (U.S. higher education, European policy, STEM fields) or broader?

Writing out your scope in 2 to 3 sentences before you start searching saves significant time. It gives you a clear criterion for deciding what to include and what to set aside.

Step 2: Search Systematically

Unsystematic searching produces uneven literature reviews that miss significant work. Use a structured approach.

Primary academic databases

Search strategy

Start with 3 to 5 core search terms related to your topic. Run each search and review the first 50 to 100 results by title and abstract. For papers that seem relevant, read the abstract carefully to confirm relevance before adding to your reference manager.

Forward and backward citation tracking is often more useful than repeated keyword searches:

• Backward: Read a key paper's references and pull the most important ones
• Forward: In Google Scholar or Web of Science, click "Cited by" on a key paper to find more recent work that built on it

This approach finds papers that use different terminology but address the same question.

How many sources do you need?

Scope determines quantity more than page count. A PhD dissertation literature review covering a specialized subfield might synthesize 80 to 150 sources. A systematic review published as a standalone article might cover 200 to 500 sources. An undergraduate thesis might review 20 to 40 sources.

Ask your advisor and look at recent published work in your area to calibrate what "comprehensive" means for your specific context.

Step 3: Read Critically, Not Passively

Reading for a literature review is different from general reading. You are not trying to absorb each paper's content. You are trying to understand each paper's relationship to your research question and to the other papers in your review.

Reading efficiently at scale: When reviewing a large pool of papers, you do not need to read every paper cover to cover before deciding whether to include it. A practical sequence:

1. Read the abstract to determine if the topic is relevant
2. Read the introduction to identify the research question and claimed contribution
3. Read the conclusion to see what the paper actually found and claimed
4. If all three confirm relevance, read the methods and results in full

This approach lets you process a large pool of candidates quickly before investing time in papers you will actually cite.

For each source you decide to include, document:

• The main argument or finding
• The methods used
• The theoretical framework
• The limitations acknowledged (or not acknowledged)
• How it relates to your own question
• Which other papers it argues with, builds on, or extends

A reference manager (Zotero, Mendeley, or similar) with a notes field for each source is almost essential for managing this at scale. Taking good notes as you read is far more efficient than re-reading papers later during writing.

Step 4: Identify Themes, Debates, and Gaps

Before writing, organize what you have found into a structure. This is the analytical step that separates a strong literature review from a weak one.

Map the themes

Group your sources by theme, theoretical approach, methodological tradition, or argument rather than by date or author. Common organizing principles include:

• By theoretical framework: How do different theoretical traditions (e.g., social constructivism vs. behaviorism, institutional vs. agent-centered approaches) approach your topic differently?
• By methodology: What have quantitative studies found versus qualitative studies? Where do the findings converge or diverge?
• By subtopic: If your topic has several distinct dimensions, organize each dimension as a section
• By debate: What are the main points of scholarly disagreement, and who stands on each side?

Identify the gaps

The most important analytical task in a literature review is identifying what has not been studied, what remains contested, or where existing methods have been insufficient. Your research question should flow directly from one of these gaps.

Common types of gaps:

• Empirical gaps: A phenomenon has not been studied in a particular context, population, or time period
• Methodological gaps: Existing studies use methods that cannot address a specific aspect of the question
• Theoretical gaps: Existing frameworks do not adequately explain observed phenomena
• Conceptual gaps: A concept is used inconsistently across the literature, creating confusion

Step 5: Write the Literature Review

Once you have your sources organized by theme and your gap identified, writing the review becomes a matter of building each section as an argument.

Opening: Scope and significance

The introduction to a literature review should:

• Define the topic and its scope
• Explain why this area of scholarship matters
• Preview the structure of the review (what themes or sections follow)

Avoid starting with "Many researchers have studied X" or "X has been studied for many years." These openings are vague and do not establish any analytical position.

Body: Theme sections

Each section of the body covers one major theme, debate, or methodological tradition. A strong body section:

• Opens with a sentence that names the theme and its significance
• Groups sources by what they argue, not by who wrote them
• Shows how sources relate to each other ("Building on Smith's finding, Jones argues...")
• Evaluates strengths and limitations of work in the section
• Ends by connecting the section to the overall gap or argument

Synthesis language is essential. Phrases that show you are synthesizing rather than summarizing:

• "Several studies find... However, this consensus is challenged by..."
• "Both Smith (2015) and Jones (2018) emphasize X, but they disagree on..."
• "While quantitative work has shown X, qualitative approaches complicate this by..."

Closing: The gap your research addresses

The conclusion of a literature review should bring all the themes together to establish the specific gap your research fills. It answers: given everything you have just reviewed, what question remains unanswered, and why is your approach to answering it the right one?

This closing paragraph is what makes the literature review function as a setup for the rest of the paper or dissertation rather than as a standalone exercise.

Common Literature Review Mistakes

Organizing chronologically. "In 1990, Researcher A said X. In 1995, Researcher B said Y. In 2000, Researcher C said Z" is a history of publication, not a literature review. Organize by argument, theme, or debate.

Summarizing without evaluating. Describing what researchers found without engaging critically with how they found it, whether it is reliable, and what it does not address is not a review. Evaluate the work, not just the findings.

Including sources to demonstrate breadth rather than relevance. Every source in your review should contribute to the argument. Sources cited only to show familiarity with the field pad the list without strengthening the review.

Failing to connect back to your own research. The literature review should build toward your research question. If a reader cannot see how your research fills a gap the review established, the review has not done its job.

Treating the literature review as complete before the dissertation is. Literature reviews are almost always revised late in the dissertation process once the final argument is clearer. The review you write in year 2 will be revised in year 4 to more precisely set up the argument your research actually produced.

Literature Reviews in Different Contexts

For PhD dissertations specifically, the literature review chapter (typically Chapter 2) is discussed in detail in the guide on how to write a PhD dissertation, including how it fits with your methods, findings, and conclusion.

Frequently Asked Questions

How long should a literature review be? Length depends on context. A PhD dissertation literature review is typically 10,000 to 20,000 words (30 to 60 pages). A journal article literature review might be 500 to 2,000 words. A standalone systematic review article may be 5,000 to 15,000 words. Ask your advisor or check examples in your field's leading journals.

Should I include everything I read? No. Include sources that contribute to the argument your review is making. Exclude sources that are tangential, methodologically too weak to be useful, or that simply repeat what better-cited work has already established. A focused 60-source review is stronger than an unfocused 150-source list.

What is the difference between a literature review and a systematic review? A systematic review follows a documented, reproducible protocol for searching and selecting sources, designed to minimize selection bias. It is a specific research method used primarily in health sciences, education, and social policy. A literature review (without "systematic") is a broader category that includes various approaches to synthesizing existing scholarship, with more flexibility in scope and method.

Can I use sources I did not read fully? You should not cite sources for specific claims if you have only read the abstract. If an article is highly relevant and frequently cited, read it fully. If it appears to support a general point that is already well established by sources you have read fully, an abstract may be sufficient for that specific use. When in doubt, read the paper.

How recent do sources need to be? For most academic fields, the bulk of your literature review should cite work from the last 10 to 15 years. Foundational, frequently cited work from earlier periods can be cited for historical context or for theoretical frameworks that have not been superseded. In rapidly changing fields (machine learning, some areas of biology), 5 years may be the practical horizon for "current."