What month should I stop fertilizing houseplants?

There is no month that works for every reader. The daylight half of our test is failed from about October to February in London, New York and Stockholm, from November in Los Angeles, in the middle of the year in Sydney, and never in Manila. That is astronomy rather than opinion, and it is why we test rather than pick a month.

When should I start fertilizing houseplants again in spring?

The same two tests, read in the other direction. The northern lines on the chart above clear the 11-hour daylight bar from about March, Sydney clears it from September, and Manila never drops below it. One thing we do not model: the common advice to skip a plant you bought or repotted in the last few months. Our growth check does not know about that, so that call stays yours.

When do houseplants go dormant?

We do not hold a dormancy date, we run a test that can flip on any day. A plant is treated as resting when the room is below the minimum for its type, or when neither the day length nor the reading at its spot clears the light bar. Two plants a couple of metres apart in one room can be on opposite sides of it in December.

Can I keep feeding through winter under a grow light?

Yes, and the bar is lower than a grow light. Any spot reading above 1,200 lux passes the light half of the test on its own, which an ordinary bright windowsill clears. Warmth still has to pass separately.

Why does GrowSpot treat my plant as resting when the room is warm?

Because warmth alone is not enough. The plant needs both a warm enough room and light, from long days or from the spot itself. In a dim corner in December neither route is open, so it is treated as resting even in a warm flat.

Does feeding pause during a heatwave?

Yes. An active heat or cold alert outranks both other tests and suppresses feeding while it runs, on the reasoning that a plant under stress cannot use it. The repot hint stands down on the same signal.

What happens if I have never taken a light reading?

The plant is treated as resting rather than growing. The app will not assume the bar has been cleared for a spot it has never seen, which is the same reason it says it cannot assess a spot rather than guessing at one.

When to Stop Fertilizing Houseplants

Stop in October, start again in March. That is the answer at the top of the results, and it is repeated by nurseries, forums and the summary above them all. It is good advice in the place it was written for.

It is also close to unanimous. Four extension services give the same window, whether they name a month or just a season, and every one of them gives it in northern-hemisphere names.

Published by	When it says to stop	When it says to start
extension.unh.edu	October	“This means putting away the fertilizer from October until the spring.” It starts again at “March is a good time to start fertilizing”.
homegarden.cahnr.uconn.edu	October	“Houseplants respond to fertilizer during periods of active growth. This is usually from March until October.”
extension.umd.edu	September	“Fertilize from March through September”, and “During the winter months, indoor plants don’t need fertilizer because reduced light and temperature result in reduced growth.”
yardandgarden.extension.iastate.edu	No month given	“Fertilization is generally not necessary during the winter months as most houseplants are not growing during this time”, and feeding runs “in spring and summer when plants are actively growing”.
extension.unh.edu	Every 1–3 months	On frequency: “most houseplants will benefit from an application once every 1 – 3 months depending on which formulation is being used.”
homegarden.cahnr.uconn.edu	Monthly	On the same question: “The labels on most water-soluble fertilizers recommend monthly applications.”

Four university extension services on the same two questions, copied exactly as each writes them. Every one of them answers in northern-hemisphere terms, whether it names a month or a season. 6 figures transcribed from 4 sources, read end to end on 30 August 2026. Units are copied exactly as each source writes them; nothing here is converted, and nothing here is our own measurement. All of them, from every page, are one free download.

GrowSpot cannot use it, because it has readers in both hemispheres and at every latitude, so the feeding decision had to be made out of things that can be measured anywhere. It ended up as two tests, and the plant has to pass both.

Worked example: Ocean Spider Plant two metres in from a north window, air 21 °C and 45% humidity, light airflow. Your spot will score differently. Run it for your own spot →

A hand tipping a metal watering can over the first of a row of small potted plants standing along a windowsill — Feeding arrives in the watering can, which is why it gets treated as part of the watering schedule. It is the one care check that reads the light instead.

The month is really a latitude

The first test is light, and the calendar-free version of it is day length: above eleven hours of daylight, the season is treated as one a plant can grow in. Day length is astronomy, so it can be worked out for anywhere from a date and a latitude, with no weather involved:

The amber line is the 11-hour bar the growth check uses. Each point is the middle of that month. Day length is pure astronomy (latitude and the date, no weather and no forecast), so this chart is the same every year. Manila never drops below the bar, Sydney crosses it in the middle of the year rather than at the end, and only the northern lines behave the way "stop feeding in October" assumes. London and Los Angeles sit between the two northern lines, clearing the bar from March to September and March to October.

Read the Manila line against the advice we started with. Near the equator the daylight test is never failed, so "stop feeding in October" is not a mild simplification there, it is the wrong answer for every month of the year. Sydney fails it in the middle of the year rather than at the end, and the two northern lines are the only ones behaving the way the advice assumes.

Indoors, the same room has two answers

Day length is not the only way to pass the light test. A reading above 1,200 lux at the spot itself passes it too, which is how a plant under a grow light stays in growth through the winter. It is also how an ordinary bright windowsill does, because that bar is lower than people expect. Feeding is the only one of the four care checks that reads light, and the watering interval does not move with it. Hold the window and move only the distance:

The amber line is the 1,200 lux bar the growth check uses. Estimates for an indoor position on a clear day, from the same model the spot checker uses. The line crosses the growth bar between 2.5 and 3 metres in. On one December day that puts a plant on the sill on the growing side of the test and one at the back wall on the resting side. A south window is a different room for this purpose: it starts at 13,000 lux and is still above the bar at the back wall, so the same question has a different answer on the other side of the flat.

That 1,200 is a threshold we score against rather than a reading we took, and it is a lower bar than it sounds: in the 204 readings behind our tiers, the indoor median sits at 2,349 lux.

The bar falls between two and a half and three metres in from a north window. That is the useful sentence on this page: on the same December day, in the same room, a plant on the sill is still growing by this test and the one by the back wall is not, and no calendar can tell them apart. Our page on how far from a window plants can live works that fall-off through in detail.

The whole test, including the part that outranks it

The second test is warmth: the average temperature has to sit above the minimum for that type of plant. And an active heat or cold alert overrides both, pausing feeding whatever the season and the sill say. Run the check itself and the shape comes out like this. A warm room with fourteen hours of daylight feeds. The same room with nine hours and 300 lux at the spot does not, and the same nine-hour day with 5,000 lux at the spot does. Twelve degrees stops it whatever the light is doing. A heat or cold alert stops it whatever everything else is doing.

The case worth pausing on is the one with no light reading taken at all. The plant is then treated as resting rather than growing, because the app will not assume a bar has been cleared on your behalf. That is the same reasoning behind the honest verdicts elsewhere on this site: an unread spot is unknown, not fine.

A windowsill of houseplants lit by several LED grow-light bars after dark, the window behind them almost black — A sill like this passes the light test in December on a reading rather than on a date, which is the whole reason the test is not a month.

What to do with your own plants

Stop feeding when the plant stops growing, and take the two measurable halves seriously: is it warm enough, and is there enough light where the plant stands. If you have a bright sill and a heated room, some of your plants may not need a winter break at all, and the ones two metres behind them will. A plant that summers outdoors clears the light half wherever it stands.

Take a reading at the spot first, with the free spot checker if you would rather not install anything. If the plant is already unhappy, feeding is rarely the fix and four things change at once in winter. If you were also planning to move it into a bigger pot, that has a season and three other conditions of its own.

The thresholds here are read out of the growth check itself, and the light figures are estimates for an indoor position on a clear day, not meter readings from a real room. The minimum temperature is held per plant type, and we do not publish those figures, because they are working defaults rather than calibrated ones. Day length is exact and identical every year.

By GrowSpot · Published August 15, 2026 · Last updated August 30, 2026

Plant photo credits