Picture your DNA match list: two hundred cousins, most of them strangers, none labelled with which side of the family they come from. The Leeds Method sorts them for you. Created by genealogist Dana Leeds, it takes your second and third cousin matches, groups them by who they share with each other, and usually produces four colour-coded columns, one for each of your grandparents’ lines.
This guide walks through the method in a spreadsheet step by step, explains which matches to include and why, shows how to read the result, and covers what to do when the columns do not come out cleanly.
Why the Method Works
Second cousins share a set of great-grandparents. Third cousins share a set of great-great-grandparents. You have four sets of great-grandparents, one couple behind each grandparent. So if you take only the matches in the second to third cousin range and group them by shared matches, they tend to fall into four clusters.
Closer matches break the logic. A first cousin shares two grandparents with you, so they would appear in two clusters at once and pull them together. That is why the method deliberately leaves them out.
What You Need Before You Start
- A DNA test with a shared-matches feature. Dana Leeds’s original walkthrough uses AncestryDNA, though the same logic works anywhere you can see which matches you share.
- A spreadsheet: Excel, Google Sheets or Numbers.
- About an hour for a typical list of 20 to 50 matches in range.
If you are unsure what a cM figure means, read our explainer on what shared cM means in a DNA match first. The method depends on choosing the right band.
The Method, Step by Step
Step 1: Pick your matches
Following the original Leeds Method walkthrough:
- Open your match list and filter to second and third cousins.
- Keep only matches sharing 90 to 400 cM.
- Remove anyone above 400 cM, and anyone you know shares two of your grandparents with you, such as a first cousin.
Step 2: List them in the spreadsheet
Put each match’s name in column A, in order of shared cM, highest first. Add column B for the cM figure. Leave columns C onwards empty for the colours. Add cell borders; it makes the colour blocks easier to read.
Step 3: Colour the first cluster
- Take match 1. Fill their cell in column C with a colour, say blue.
- Open match 1’s shared matches. For every shared match who is on your list, fill their column C cell blue too.
Step 4: Start the next column
- Go down your list to the first match with no colour.
- Give them a new colour in column D, say orange.
- Open their shared matches and colour each one orange in column D. If someone is already blue, colour them orange as well. Do not skip them.
Step 5: Repeat until everyone has a colour
Keep moving to the next uncoloured match, starting a new column each time, until every match in your list has at least one colour.
Copying names in from the website is the slow part. Paste them as plain text and use a fill-down or format-painter shortcut to colour quickly; this list of everyday keyboard shortcuts covers the copy, paste and undo combinations that speed up a long session.
A Small Worked Example
Here is how a short, invented list might look after sorting. Each letter stands for a colour column.
| Match | cM | Column |
|---|---|---|
| Match 1 | 310 | Blue |
| Match 2 | 244 | Orange |
| Match 3 | 201 | Blue |
| Match 4 | 178 | Green |
| Match 5 | 150 | Orange and Green |
| Match 6 | 122 | Purple |
| Match 7 | 96 | Green |
Four colours: a promising result. Match 5 sits in two columns, which needs a closer look (see below). If you know your mother’s cousin Match 1 descends from her father’s side, blue becomes “maternal grandfather”, and every blue match is a lead on that line.
Reading Your Columns
| You see | Likely meaning |
|---|---|
| Four columns, little overlap | The four great-grandparent couples behind your grandparents |
| Three columns | No tested cousins in range from one line yet |
| Four columns with some overlap | Two clusters may be on the same side, or one line is missing |
| Many columns, heavy overlap | Possible pedigree collapse or endogamy |
Treat every column as a hypothesis. The colours point you towards a grandparent line; documents and family trees confirm it.
When a Match Sits in Two Columns
Overlap is normal, and it is a clue rather than an error. Check these possibilities in order:
- Two clusters belong to the same grandparent. For example, one column could be your grandmother’s father’s family and another her mother’s family. A match related to both joins them.
- The match is related to you twice. In small communities a cousin can be connected through two lines.
- A coding slip. Recheck that match’s shared matches; Dana Leeds advises confirming before you accept the overlap.
If the same two colours overlap again and again, they are probably one side of the tree split in two.
When the Method Struggles
- Endogamy. Where a community married within itself for generations (for example some Jewish, Amish, Acadian or island populations), everyone shares with everyone and the columns blur. The method is not designed for this.
- Pedigree collapse. If cousins married in your own recent tree, expect heavy overlap.
- Few matches. With only a handful of matches in range, one line may not show up at all.
- Small testing families. If one grandparent came from a country where few people test, that column may stay empty for years.
The fix for thin results is usually more testers. The oldest living relatives give the most useful matches. If you are choosing a kit for them, our comparison of which DNA test is best for genealogy sets out database sizes and privacy settings.
Turning Columns Into Ancestors
- Label each column with the line it represents once you have one confirmed match.
- Look at the trees of matches in each column and note surnames and places that repeat.
- Build a quick tree for any untreed match from public records; the common couple usually emerges after two or three trees.
- Add the colours as tags or groups on the testing site, so the sort follows you when you log in.
Run the method once on a quiet evening, label what you can, and revisit it every few months as new cousins test. The spreadsheet grows more useful each time.
Frequently Asked Questions
What cM range does the Leeds Method use?
The original walkthrough uses matches sharing 90 to 400 cM, which roughly covers second and third cousins. Matches above 400 cM are left out because a close relative such as a first cousin shares two grandparents with you and would merge two columns. Some people widen the lower limit when they have few matches.
Does the Leeds Method work outside AncestryDNA?
Yes, in principle. The method only needs a list of matches and a way to see which matches you share with each one. MyHeritage, FamilyTreeDNA and others offer shared-match lists. The method was described using AncestryDNA, so the filters and labels you see elsewhere will look different.
Why do I only get three columns?
Most often because no one from one grandparent’s line has tested yet, or because their matches fall below 90 cM. It can also happen when a grandparent was an only child from a small family. Encourage relatives on the missing side to test, or try lowering the cM floor carefully.