Coffeeberry: The Shrub That Was Never Coffee

A field guide to Frangula californica, the evergreen shrub that does almost everything a Bay Area garden asks of it and still gets passed over at the nursery, carrying a common name that gets nearly everything about it wrong. What it actually is, what it wants from you, and why a band-tailed pigeon in October has a better claim on the berries than you do.

About the Plant

Coffeeberry is the plant that should be an easy sell and somehow isn't. It's evergreen, so it holds the bed through winter. It's long-lived past the point of absurdity. It takes sun or part shade, clay or serpentine, and it's one of the few California natives that won't sulk if your sprinklers reach it. It's widely enough grown that the Forest Service calls it frequently cultivated ornamentally (McMurray, 1990), and there are at least six named selections in the trade, from a two-foot groundcover to a small tree.

And the California Native Plant Society still describes it as often overlooked as a garden subject (Mazur, 2025). Which is about right. Nobody walks into a nursery in April looking for a shrub with greenish flowers you can barely see.

The name doesn't help. There's no coffee in it and no caffeine either, and the family it belongs to made its money for about a century on a bark that emptied people out.

None of that is a reason to skip the plant. It's a reason to stop taking the name at face value. Underneath it is a shrub that will very likely outlive you, feeds birds in the leanest month of the California year, and won't hold a grudge about your irrigation, which puts it in a very small club.

What You're Looking At

Start with the leaf, because that's what you'll see first and most of the year. Coffeeberry is evergreen, with elliptical leaves running from under an inch to about four inches long (2–10 cm), glossy on top. Young twigs are usually reddish; older branches go gray or brown (McMurray, 1990; Sawyer & Nesom, 2016). The leaf undersides vary a lot depending on which subspecies you're looking at: bright green, yellowish, or in the hoary types densely covered in fine white star-shaped hairs (Sawyer & Nesom, 2016).

Leaf margins on this plant are typically inrolled to begin with, meaning the edges curl under toward the leaf's underside rather than lying flat (McMurray, 1990), and they roll further in a dry summer, which is the plant conserving moisture rather than the plant in trouble (Golden Gate National Parks Conservancy, n.d.). Leaves also run small and thick on dry sites and larger and thinner in moist ones (McMurray, 1990), so two coffeeberries fifty feet apart can look like different plants.

Flowers come April into June and you have to go looking for them. They're small and yellow-green, five petals each, gathered in stalked umbels tucked into the leaf axils (Tutka, 2015; Sawyer & Nesom, 2016). Nothing about them says "look here." The bees have a different opinion.

The fruit is what the plant is named for, and it's the part worth looking at closely. It's a juicy drupe (a fleshy fruit built around a stony pit), running from pea-sized up to about half an inch (7–15 mm), and it moves from green through dull red to glossy black as it ripens (McMurray, 1990; Tutka, 2015; Sawyer & Nesom, 2016). Ripening runs across a long stretch, roughly July through November, with dispersal in fall (McMurray, 1990), which is why a shrub in September can look like it can't make up its mind about what color it wants to be. Inside each fruit sit two seeds, sometimes three depending on the subspecies (Sawyer & Nesom, 2016). Split one open. That's the "coffee" bean.

The Name Doesn't Survive Contact with the Plant

Three things are wrong with "coffeeberry," and they get worse as you go.

The first is the smallest. It's a resemblance and nothing more, and the sources can't even agree on which part of the plant is doing the resembling. The Forest Service species review says the two seeds inside each fruit closely resemble the commercial coffee bean (McMurray, 1990). The California Native Plant Society credits the appearance of the ripe berries instead, and takes care to add that the name has nothing to do with edibility (Mazur, 2025). Split a fruit open in October and judge for yourself. Nobody in either camp is claiming it tastes like anything.

The second one is funnier, because people tried. Somebody looked at a shrub loaded with coffee-bean lookalikes, in a state where coffee doesn't grow, and did the obvious experiment. It didn't work. Attempts at using coffeeberry as a coffee substitute have not been successful, and that one-line verdict has stood in the federal record, unelaborated, for decades (McMurray, 1990). Nobody wrote down what it tasted like. Whatever happened was apparently enough.

The third is chemical. Caffeine and its relatives are purine alkaloids (a small group of nitrogen-ring compounds built on a shared chemical skeleton), and the plants known to make them are a scattered handful of lineages: tea, coffee, cacao, and holly among them (Ashihara et al., 2008). Coffeeberry's family, Rhamnaceae, is not one of the caffeine-accumulating groups described in that literature. It's also nowhere near real coffee on the family tree: coffeeberry is a buckthorn (Tutka, 2015), and Coffea is in an entirely separate branch of the flowering plants. The resemblance stops at the seed coat.

The plant's own name has been unstable too. It went into the literature in 1826 as Rhamnus californica and is now Frangula californica (Sawyer & Nesom, 2016), a change that followed the use of nuclear and chloroplast DNA sequences to redraw the genus-level boundaries inside Rhamnus in the broad sense, the old catch-all version of the genus (Bolmgren & Oxelman, 2004). Most nurseries still print the old name. Both get you the right plant.

Thin Cascara Bark

The family history is where a harmless misnomer stops being harmless.

Coffeeberry's close relative is Frangula purshiana, cascara sagrada, Spanish for sacred bark. Its bark carries anthraquinone compounds and their glucosides, the cascarosides, which act locally as an irritant on the colon and force it to contract (National Institute of Diabetes and Digestive and Kidney Diseases, 2017). That made it one of the biggest botanical drugs in North America, and the harvest ran at industrial scale: in a single year, five million pounds of dried Pacific Northwest bark went to pharmaceutical manufacturers (Habeck, 1992).

A trade that size creates a supply problem, and supply problems create substitution. Coffeeberry bark looks very much like cascara bark, only thinner. By 1898, California druggists had a trade name for it: thin cascara bark. King's American Dispensatory, the standard reference of the era, records the practice second-hand, quoting a contemporary who wrote that it had doubtless been a common practice to sophisticate cascara with the bark of Rhamnus californica (Felter & Lloyd, 1898). "Sophisticate" was the polite pharmacy word for adulterate. A generation later the same observation was still being made, and by then coffeeberry had earned its own entry in the materia medica (the era's working catalogue of drug plants), with dosing instructions and a listed action as a cathartic, meaning a laxative (Felter, 1922).

So a plant now sold as a tidy foundation shrub spent part of its career as a cut agent in the patent-medicine trade.

Cascara itself stayed on American pharmacy shelves for another century. Then in May 2002 the Food and Drug Administration (FDA) issued a final rule stating that cascara sagrada and aloe were not generally recognized as safe and effective as over-the-counter (OTC) laxative ingredients, and directed that they be removed from OTC laxative products within 180 days (67 Fed. Reg. 31125, 2002). The rule turned on missing safety data rather than on proof of harm; manufacturers had been asked for carcinogenicity and genotoxicity (cancer-causing and DNA-damaging) studies and hadn't produced them. Either way, the sacred bark came off the shelf.

Where It Comes From

Coffeeberry runs from extreme southwestern Oregon south along the coast and Coast Ranges into southwestern California (McMurray, 1990), with the species as a whole reaching into Arizona, New Mexico, Nevada and northwestern Mexico (Sawyer & Nesom, 2016). Around here it's below about 5,500 ft (1,677 m), on dry flats, moist slopes, ravines and rocky ridges (McMurray, 1990). It turns up in chaparral, coastal sage scrub, mixed-evergreen forest, redwood forest and oak woodland (Tutka, 2015), which is already a hint about how little it cares where you put it.

There are six recognized subspecies, and Flora of North America is refreshingly honest about how well they hold up: they're more or less separated geographically, and intermediates exist among all of them except one (Sawyer & Nesom, 2016). Bay Area plants are subspecies californica, whose range runs from southern Oregon down to Orange County (Tutka, 2015).

The Mistake Most People Make

Somebody buys a one-gallon coffeeberry with no cultivar name on the tag, puts it in a four-foot gap between a path and a fence, and finds out four years later what they actually bought.

Frangula californica grown from unselected seed is typically a 4 to 6 ft (1.2–1.8 m) shrub in the wild and occasionally becomes a small tree of about 20 ft (6 m). On a suitable site, transplants can reach 8 to 12 ft (2.4–3.7 m) with comparable spread inside about twenty years, and irrigated sites throw large numbers of self-sown seedlings (McMurray, 1990). Nursery listings for the species land anywhere between 4 and 20 ft depending on who's writing them, which is a long way of saying that unnamed seedling stock is a coin flip.

The named selections exist to solve exactly this, and there's a specific history behind them. The Saratoga Horticultural Research Foundation started introducing California natives to the nursery trade in the early 1950s, when natives were widely regarded as weeds and almost nothing was commercially available; over fifty-two years it released fifty cultivars of native and non-native plants (Coate & Coate, 2024). Coffeeberry got several of them, and the useful ones for a Bay Area garden are:

• 'Eve Case' is the standard. A Saratoga Horticultural Foundation introduction from 1975, most often 4 to 6 ft but capable of a mound 8 to 10 ft by as wide, with showy berries (San Marcos Growers, n.d.-a). If someone hands you a coffeeberry and doesn't say which one, it's probably this.
• 'Mound San Bruno' was selected by Roger Raiche from plants on San Bruno Mountain in San Mateo County. Roughly 4 to 6 ft tall and about twice as wide as tall, with foliage around half the size of typical coffeeberry leaves (San Marcos Growers, n.d.-c). The one to use when you want width without height.
• 'Seaview' is the smallest thing in the category, about 2 ft tall by 4 ft wide, introduced by the same Saratoga foundation and useful as an edging plant (Yerba Buena Nursery, n.d.). Effectively a groundcover.
• 'Leatherleaf' came out of the UC Berkeley Botanic Garden, from a selection Roger Raiche made on Montara Mountain. Darker leaves than the other cultivars, with margins rolled under for a heavy, almost rhododendron-like look. Usually 5 to 6 ft, occasionally 8 (San Marcos Growers, n.d.-b).

Buy by name. The tag is the only thing standing between you and a shrub three times the size of the space you have.

A fair question if you garden near wild stands, meaning naturally occurring populations rather than planted ones: does buying a selection instead of the species itself actually matter? Nobody has studied gene flow in coffeeberry specifically, so anyone who gives you a clean answer is extrapolating. The general native-cultivar literature is clear that cultivars carry narrow genetic diversity produced by artificial selection, and that "plant it far enough away" is not a real mitigation: cultivar pollen has been tracked as far as 12.4 miles (20 km) into wild populations, in work on wind-pollinated bentgrass (Tangren, 2019). Coffeeberry is insect-pollinated, so that distance almost certainly overstates its case. In a developed neighborhood, plant the selection and enjoy it. If you're doing restoration work next to intact chaparral, source local-provenance species stock, meaning seed collected from wild populations near your site.

Sun: More Latitude Than Almost Anything Else Here

Coffeeberry takes full sun through moderate shade once established, and the plant changes shape depending on which it gets. In sun it stays compact, with lighter grey-green foliage. In shade it grows taller and more open, with darker leaves, and can reach fifteen feet or more (Mazur, 2025). Both are the plant behaving normally.

The nursery trade rounds this up to "sun or full shade." The sourced version doesn't go that far. The Forest Service species review describes coffeeberry as moderately shade-tolerant and caps the range at moderate shade (McMurray, 1990). In practice that means a north-facing bed with open sky above it will do fine, and the dark strip under a mature oak on the north side of a fence is asking more than the plant reliably delivers. If you need something for genuinely deep shade, last week's Douglas iris entry is the better answer.

Water: The Native That Doesn't Hold a Grudge

Most of this series has been a variation on one warning: turn the hose off or lose the plant. Coffeeberry is one of the few entries where that warning doesn't apply, and it's worth saying plainly, because by now the reflex is strong.

Established plants need no irrigation at all (Oregon State University Department of Horticulture, n.d.). That part is standard. What's unusual is the other direction: coffeeberry tolerates sand, clay, and even occasional seasonal flooding (Tutka, 2015), and the Forest Service review notes that large numbers of volunteers, meaning self-sown seedlings, turn up on sites receiving additional irrigation (McMurray, 1990). Give this plant extra water and it makes more coffeeberries. It doesn't rot out in its second winter.

That single trait is why it's the most useful native in this series for anyone converting a yard in stages. You can put a coffeeberry into an existing irrigated bed, next to whatever is already there, without redesigning your watering first. Very little else in this pillar will accept that deal.

New plants still want water through their first dry summer while the roots get out into the surrounding soil. After that it's optional, and the plant will tell you nothing either way.

Soil: It Genuinely Doesn't Care

Wild coffeeberry grows on soils that are typically dry and well drained (McMurray, 1990), but the tolerance range is much wider than that. It handles sand and clay (Tutka, 2015), and it grows on serpentine-derived ground, the soils that run high in magnesium, chromium and nickel and low in calcium and potassium, which is why most plants want nothing to do with them. Coffeeberry associations occur on ultrabasic soils in the Siskiyou Mountains (McMurray, 1990), and subspecies occidentalis characteristically grows on mafic and ultramafic substrates (Tutka, 2015), the iron- and magnesium-rich rock family serpentine belongs to. If you garden on one of the serpentine outcrops scattered around Santa Clara County, this is a shortlist plant.

It doesn't need soil amendment, and there's no drainage engineering to do. Dig the hole, backfill with what came out, water it in.

What Happens After a Fire

Coffeeberry is a root-crown sprouter, and the detail is more specific than that phrase usually implies.

After fire or cutting kills the top growth, the plant sprouts vigorously from dormant buds on the root crown, which stores both buds and carbohydrates and lets the shrub reoccupy the site fast (McMurray, 1990). Most plants survive fire even when the entire above-ground structure is gone, and they regain their prefire size and biomass quickly. The root crown enlarges with repeated sprouting, but it isn't a lignotuber, the specialized woody swelling that manzanita and some other chaparral shrubs carry. It's an ordinary root crown that has been through a lot.

Seedlings are a murkier story, and the Forest Service review says so outright: the pattern of postfire seedling establishment in coffeeberry is unclear. In southern California chaparral it apparently behaves as an obligate sprouter, meaning it returns only from the surviving root crown rather than from seed, and rarely establishes seedlings after a burn, while on northern California chaparral sites one study found 4,400 seedlings per acre on one-year-old burns (McMurray, 1990). What stays consistent is the seed itself, which is short-lived, holding viability no longer than about nine months when dried at room temperature, so there's no long-term bank of it sitting in the soil waiting for a fire (Keeley, 1987, as cited in McMurray, 1990). The root crown is the part you can count on.

The practical translation for a garden: coffeeberry tolerates pruning well, regenerates from the root crown, and makes a decent informal hedge if you want one (Tutka, 2015; Mazur, 2025). Nothing you do with loppers is worse than what it evolved to survive.

The Berries, and Who They're For

The most useful line anyone has written about coffeeberry fruit comes secondhand through the Forest Service review, from field notes at the Hastings Natural History Reservation in Monterey County: in many areas, the berries are often the only abundant juicy fruit available in the fall (Griffin, 1974, as cited in McMurray, 1990).

That's the whole ecological argument for the plant. Autumn in a California landscape is a lean stretch. Coffeeberry is carrying soft fruit through it, and bird harvest is so complete that relatively few seeds ever fall beneath the parent plant. Band-tailed pigeons are the species most consistently named in the literature; black-tailed deer and black bear take the fruit too, and woodrats eat the seeds in small quantities. Shrubs with bird-dispersed seed, coffeeberry among them, have apparently increased on relic oak savanna sites in central California (McMurray, 1990), which is to say the dispersal system works well enough to move a landscape around.

Which brings us back to the name, and what you should do about it.

One Honest Caveat: Don't Make Anything Out of It

The Forest Service species review says the berries are sweet and edible and were gathered historically for culinary as well as medicinal purposes (McMurray, 1990). That's the only affirmative statement on edibility I found in a reputable source. San Marcos Growers, one of the largest wholesale growers of this plant in the state, lists it the other way, as possibly poisonous (San Marcos Growers, n.d.-a).

I would leave the berries alone.

The direction of this plant's chemistry isn't in dispute either way. Coffeeberry produces a purgative effect, and that effect is the documented basis of its traditional use. The plant guide from the U.S. Department of Agriculture's Natural Resources Conservation Service (NRCS) records that Chumash and Costanoan people dried and ground the inner bark to make a laxative tea, along with other medicinal preparations from the leaves and bark (Tutka, 2015). Costanoan is the older ethnographic term for the Ohlone.

So the practical position is easy. The berries are for birds. Don't make anything out of the bark. The seeds aren't coffee, they contain no caffeine, and the one recorded verdict on trying them is that it didn't work. Its closest relative's bark was a mainstream pharmaceutical for a century and got pulled from over-the-counter sale in 2002 for want of safety data. That is not a lineage to freelance with in your kitchen.

Plant it and look at it. The pigeons get the rest.

Sourcing

Buy coffeeberry in fall from a California Native Plant Society (CNPS) chapter sale or a native-plant nursery, and buy it by two names: the Latin binomial and the selection.

Ask for Frangula californica and expect some nurseries to have it filed under Rhamnus californica, which is the same plant under its older name. Then ask which selection it is. Given a species that runs from a two-foot groundcover selection up to a twenty-foot tree, "coffeeberry" on a tag is not enough information to site the plant. If nobody at the nursery can tell you, treat it as unselected stock and give it room for 8 to 12 ft.

Plant fall through winter, the same window that governs everything else in the California Natives pillar, so the roots get a wet season before the first dry summer. Water through year one. After that the plant is on its own schedule, and its schedule runs to a century or two (McMurray, 1990).

The Takeaway

Coffeeberry is the plant you put in when you want the question settled. Sun or part shade, sand or clay or serpentine, irrigated or bone dry, and it holds its leaves through winter regardless. Buy a named selection so you know what size you're getting. Put it in the ground in fall, water it through the first dry summer, and then leave it alone for the next hundred years. The berries were never for you anyway. Sometime in autumn a band-tailed pigeon works out that your shrub is one of the few things in the neighborhood still carrying juicy fruit that late in the year, and the crop goes fast. That's what you bought it for.