Affiliate disclosure: some links on this page go to Amazon. If you buy through one we may earn a small commission, at no extra cost to you. Full disclosure.
Project FeederWatch is a winter-long bird survey run by the Cornell Lab of Ornithology and Birds Canada, running from November to April. You count the birds at your feeders over two consecutive days, as often or as rarely as you like across the season, and submit the results. There is a small annual participation fee, and no experience is required.
If the Great Backyard Bird Count is a weekend, Project FeederWatch is the season. It has been running since the 1980s, and the resulting dataset is the reason a great deal of what you read about American feeder birds — which seeds species prefer, how ranges are shifting, how feeder-borne disease spreads — is known at all.
Much of the guidance on this site cites FeederWatch. Taking part is how that guidance gets written.
The essentials
- Season: November through April
- Commitment: as little as one two-day count in the whole season
- Cost: a small annual participation fee, which funds the project
- Where: your own feeders, at home
- Run by: Cornell Lab of Ornithology and Birds Canada
- Experience needed: none
How the counting works
The protocol is specific, and following it is what makes the data usable. It is not complicated.
1. Define your count site. The area you can see from one spot, containing your feeders. It stays the same all season.
2. Count over two consecutive days. A count period is a pair of days. You can watch for five minutes or several hours across those two days.
3. Record the highest number seen at once. The same rule as the GBBC and the same reason: it prevents counting one individual repeatedly. If you see four juncos together at any point, the answer for that count is four, however many times they come and go.
4. Leave at least five days between counts. Then start a new count period whenever you want.
5. Submit online, through the FeederWatch site or app.
The reporting of nothing matters too. Weeks when you watched and saw very little are as valuable as busy ones, because absence is data. Skipping a submission because “there was nothing worth reporting” removes exactly the signal that shows a species declining or a disease outbreak passing through.
What comes out of it
Four decades of standardized counts from thousands of yards produce things a research team could not collect any other way:
- Seed and feeder preference data — the basis for most of what is known about what species actually take
- Irruption tracking — Pine Siskin, redpoll and Evening Grosbeak movements south in poor cone years
- Range shifts, including species expanding north over decades
- The House Finch Disease Survey, which has tracked mycoplasmal conjunctivitis since it emerged in House Finches in the early 1990s — one of the clearest examples of citizen science documenting a wildlife epidemic in real time
- Salmonellosis outbreak detection, including the large Pine Siskin die-off of 2020–21
Participants are also asked to report sick birds, which feeds directly into that disease work. If you see a finch with swollen or crusted eyes, reporting it is a genuine contribution — see sick birds at your feeder for what to do about the feeders themselves.
The fee
Project FeederWatch charges a modest annual participation fee, unlike the free GBBC. It funds the staff who validate submissions, the data infrastructure, and the materials sent to participants. Cornell Lab members generally participate at no additional charge, and Birds Canada handles Canadian participation separately.
Fees change, so check the official site rather than any figure quoted elsewhere.
Setting up for it
Nothing special is required, but a few things make a season of counting more productive:
- More than one feeder type. A tube and a platform, or a tube and a suet cage, produce a noticeably longer species list than one feeder alone
- Consistency. A reliably stocked feeder is visited more consistently, and consistency is what a season-long survey is measuring
- Water, which brings in species that never use a feeder
- Cleanliness. Counting birds all winter at a feeder you are not cleaning every two weeks is working against the project’s own findings
- A comfortable chair by the right window. Not a joke — the counts you actually do are the ones that are easy to do
If you are adding a feeder before the season, a tube with metal-reinforced ports is the useful all-rounder — all-plastic ports get chewed wide by squirrels within a season or two. The drawback of tube feeders generally: seed below the lowest port sits and can go moldy, so empty that last inch at each cleaning rather than topping up over it. Full comparison in types of bird feeders.
Which project to choose
| If you want | Try |
|---|---|
| To try it once, for free | Great Backyard Bird Count, four days in February |
| A regular winter routine at home | Project FeederWatch |
| A full day out with other people | Christmas Bird Count |
| To follow nests through spring | NestWatch |
| To record birds anywhere, any time | eBird |
Common questions
When does Project FeederWatch run?
November through April each year. You can join partway through a season.
How much time does it take?
As little as you want. One two-day count in an entire season is a valid contribution, and there is no minimum watching time within those two days.
Does it cost anything?
There is a small annual participation fee that funds the project. Cornell Lab members generally take part at no extra charge. Check the official site for the current amount.
How do I count birds that keep coming and going?
Record the highest number of each species visible at one moment during the count period. This prevents counting the same individual repeatedly.
Should I submit a count where I saw almost nothing?
Yes. Low counts and absences are as informative as busy ones, and dropping them biases the dataset toward good weeks.
Do I need to identify every bird?
No. Report what you are confident about. Merlin Bird ID from the Cornell Lab helps with the rest, and FeederWatch provides regional identification materials.
Related guides
- The Great Backyard Bird Count
- Sick birds at your feeder
- Common backyard birds of North America
- Types of bird feeders
Sources: Project FeederWatch, Cornell Lab of Ornithology and Birds Canada, including the count protocol and the House Finch Disease Survey; Cornell Lab of Ornithology, All About Birds, eBird and Merlin Bird ID; National Audubon Society.
Starting out?
One tube feeder with black oil sunflower brings in more species than any other single setup. Metal ports last; all-plastic ports get chewed open within a season or two.
This link goes to Amazon. If you buy through it we may earn a small commission, at no extra cost to you.