The short version
A logic model is a one-page picture of how your program is supposed to work: the resources you invest (inputs), what you do (activities), what you produce (outputs), and the changes that follow in the short, intermediate, and long term (outcomes). A theory of change explains why those links should hold. Build the logic model before you write the narrative, and the rest of the proposal gets easier.
A logic model is the program on one page. It shows, column by column, how what you invest and do leads to the changes you promise. Reviewers use it to check whether your plan hangs together. You should use it for something more valuable: to force the program, finance, and evaluation decisions before anyone writes a paragraph.
CDC's 2024 Program Evaluation Framework describes a logic model as a graphic depiction of the relationship between a program's activities and its intended outcomes, showing the sequence of events for bringing about change and the "if-then" relationships between activities and outcomes. The Administration for Community Living's guidance describes it as a visual, systematic way to show how resources, activities, and anticipated results connect.
The components
| Component | What it is | Examples | Common mistake |
|---|---|---|---|
| Situation / problem | The need the program addresses | Third graders reading below grade level | Restating your program instead of the problem |
| Inputs | Resources invested | Staff, volunteers, funding, curriculum, space, partners, data systems | Listing only the grant money |
| Activities | What the program does with inputs | Recruit and train tutors; deliver tutoring; host family workshops | Too granular (every task) or too vague ("provide services") |
| Outputs | Direct, countable products of activities | 360 sessions delivered; 120 students enrolled; 12 workshops held | Writing outcomes here |
| Short-term outcomes | Early changes in knowledge, skills, attitudes, awareness | Students master target phonics skills | Promising behavior change too early |
| Intermediate outcomes | Changes in behavior, practice, or performance | Students read independently; reading level rises | Skipping this column |
| Long-term outcomes (impact) | Changes in conditions or status | Students read proficiently by fourth grade | Claiming long-term impact within one grant year |
| Assumptions | Beliefs that must be true for the links to hold | Students will attend regularly; schools will share assessment data | Leaving them unstated |
| External factors | Conditions outside your control that affect results | School staffing changes; family mobility; policy changes | Ignoring them |
The ACL guidance follows the same sequence (need, inputs, activities, outputs, immediate and intermediate outcomes, ultimate outcomes) and recommends documenting external factors and assumptions.
Theory of change vs. logic model
The terms are often used interchangeably, but there is a useful distinction.
| Logic model | Theory of change | |
|---|---|---|
| Question it answers | What will we do and what will result? | Why and how will change happen? |
| Format | Columns or a flow diagram, one page | Narrative, pathway map, or diagram with causal links |
| Focus | One program's operations and results | The causal pathway and assumptions, often beyond one program |
| Detail on assumptions | Usually brief | Central: explicit about each causal link and the evidence for it |
| Typical use in grants | Required attachment; basis for evaluation plan | Narrative section; strategy documents; larger initiatives |
CDC's framework notes that a program description should clarify the program's hypothesized mechanism, or theory of change, explaining how activities are presumed to lead to the intended outcomes, with any supporting evidence cited and evidence gaps identified. In other words: the logic model shows the chain, and the theory of change justifies each link.
Federal definitions vary. The U.S. Department of Education, for example, defines a logic model as "also referred to as a theory of action": a framework that identifies the key project components (the "active ingredients") and describes the relationships among those components and relevant outcomes (34 CFR 77.1). Always use the funder's definition and format when they give one.
Logic models and evidence requirements
Some federal programs tie logic models to evidence standards. Under the Department of Education's definitions in 34 CFR 77.1, the lowest evidence level, "demonstrates a rationale," means a key project component in the project's logic model is supported by citations of high-quality research or evaluation findings suggesting it is likely to improve relevant outcomes. If a NOFO references these evidence levels, your logic model is part of how you meet them. See program design for more on evidence.
A full worked example
Situation: At three Eastside elementary schools, 58% of third graders read below proficient (vs. 39% districtwide). No structured after-school reading support exists; 210 students in grades 2–4 need intensive help.
Goal: Every Eastside child reads proficiently by the end of third grade.
| Inputs | Activities | Outputs | Short-term outcomes (Oct–Jan) | Intermediate outcomes (by May) | Long-term outcomes (2–4 years) |
|---|---|---|---|---|---|
| 1 FTE program director (reading specialist) | Recruit and screen students using fall benchmark scores | 120 students enrolled (40 per school) | Students master grade-level phonics and decoding skills | 70% of students attending 60+ sessions gain at least one grade level in reading | Higher share of participants score proficient on the state third-grade reading assessment |
| 3 part-time site coordinators | Recruit, train, and coach 9 part-time tutors | 9 tutors complete 20 hours of training; monthly coaching observations | Students read aloud more fluently (words correct per minute) | Students read independently at home three or more times a week | Narrowed gap between Eastside and district reading proficiency |
| 9 part-time tutors | Deliver four 75-minute small-group sessions per week, Oct–May | 360 sessions delivered (120 per school); average attendance 80% | Students report greater confidence and enjoyment in reading | Teachers report improved classroom reading participation | Improved school engagement and grade-level progress for participants |
| Structured phonics-and-fluency curriculum | Monthly family literacy workshops | 12 workshops; average 25 parents | Parents learn three or more strategies to support reading at home | Families read together more often | Stronger home literacy environment |
| Classroom space at 3 schools (in-kind) | Share progress reports with teachers and families each quarter | 3 progress reports per student | Teachers and families aware of each student's progress | Coordinated support between school and program | |
| District assessment data-sharing agreement; external evaluator; $85,000 grant plus $27,000 match | Collect attendance and fidelity data; pre/post assessments | Complete attendance records; fall and spring scores for all students |
Assumptions: Students will attend at least 60 sessions if programming is on-site right after school. Small-group, structured phonics instruction improves decoding and fluency for struggling readers (cite the evidence you rely on). Schools will continue to share benchmark data under the agreement.
External factors: Teacher turnover at partner schools; family moves during the year; changes in the district's benchmark assessment; weather and school closures.
Notice that each outcome is something the evaluation plan can measure, and each activity appears with a cost in the budget. The numbers (three schools, 120 students, 360 sessions) match the objectives exactly.
Download the logic model template to build your own.
How to build a logic model, step by step
The ACL guidance recommends starting with the need, then working through inputs, activities, outputs, and outcomes. Many practitioners find it easier to start at the right-hand side. Either works; here is the outcome-first approach.
- Write the problem and long-term goal. Pull them from your statement of need.
- Work backward from long-term outcomes. Ask: what must change in the intermediate term for that to happen? Then: what must change first (short-term)?
- Identify the activities that produce those short-term changes. Cluster them into four to six groups.
- List outputs for each activity: what you will count.
- List inputs: every resource the activities need, including in-kind and partner contributions.
- Write assumptions and external factors.
- Test every link with "if-then" logic, and ask colleagues, partners, or participants to challenge it.
- Check alignment with objectives, timeline, budget, and evaluation.
- Simplify to one page.
Writing the narrative that accompanies the model
Most funders who ask for a logic model expect a short narrative alongside it, or they expect the project narrative to reflect it. CDC's framework describes the program roadmap as typically a one-page graphic accompanied by a narrative explanation with more detail. A good accompanying narrative does three things: walks the reader through the main causal pathway, states the key assumptions and the evidence behind them, and flags which outcomes the evaluation will measure during the grant period.
Using the logic model after the award
A logic model is not only a proposal attachment. Once funded, it becomes a management tool:
- Staff orientation. New staff and volunteers understand quickly how their work connects to results.
- Monitoring. Review outputs monthly against the targets in the outputs column. If outputs fall short, outcomes will too.
- Course correction. When results disappoint, the model helps you find the broken link: was it low attendance (an output problem), or did attendance not translate into skill gains (an outcome problem)?
- Reporting. Structure grant reports around the model's columns so funders see progress against the plan they approved.
- Future proposals. A model refined with real data makes your next proposal stronger and your targets more defensible.
Formats
- Table (columns): Most common, easiest to read, and what most funder templates use.
- Flow diagram: Boxes and arrows, useful when activities lead to different outcomes. Keep arrows meaningful; a diagram where everything connects to everything says nothing.
- Funder template: If the funder provides one, use it exactly, even if you prefer a different format.
Common mistakes
- Outputs in the outcomes column. "120 students served" is an output. Outcomes describe change.
- Missing intermediate outcomes. Jumping from "attended sessions" to "graduated high school" skips the links a reviewer needs to believe.
- Activities that produce nothing in the outcome columns, or outcomes no activity could produce.
- Outcomes you won't measure. Mark which outcomes your evaluation will track, especially long-term ones you can't measure in the grant period.
- Numbers that don't match the narrative.
- Too much detail. If it doesn't fit on a page, you are writing a work plan, not a logic model.
- Treating it as a one-time attachment. Revisit it when the program changes and use it to frame your grant reports.
For how the logic model fits with every other section, see the anatomy of a grant proposal.
Common questions
What are the components of a logic model?
The standard columns are inputs (resources), activities (what the program does), outputs (direct, countable products of activities), and outcomes, usually split into short-term, intermediate, and long-term. Many logic models also include the problem or situation, assumptions, and external factors that could affect results. CDC's evaluation framework notes the most common elements are activities and outcomes, sometimes with inputs.
What is the difference between a logic model and a theory of change?
A logic model shows what the program does and what results it expects, in a structured table or diagram. A theory of change explains why and how change is expected to happen, including the assumptions and causal pathways behind each link, and often looks beyond one program. In practice, many funders use the terms loosely; read the guidelines to see which they want.
Do all grant applications require a logic model?
No. Many foundations don't ask for one, but most federal agencies and many larger foundations do, and some U.S. Department of Education programs use the logic model as part of how they define evidence. Even when not required, building one helps you write consistent objectives, program design, and evaluation sections.
How long should a logic model be?
One page is the convention, in landscape orientation. If you need more space, your model is probably too detailed: summarize activities into four to six clusters and focus outcomes on the ones you will actually measure. Put detail in the program design and evaluation narrative instead.
Sources
We check facts against primary sources wherever possible. Rules and programs change, so confirm details with the funder or agency before you apply.
- CDC Program Evaluation Framework, 2024 — MMWR Recommendations and Reports
- ACL's Logic Model Guidance — Administration for Community Living
- 34 CFR 77.1 Definitions that apply to all Department programs (logic model; demonstrates a rationale) — eCFR
Spotted something out of date? Send a correction.