414 guides across 21 subjects
Every figure traced to a named source
Prices in Ugandan shillings, per acre and per animal

Robusta vs Arabica Coffee Farming

A coffee bush carrying ripening cherries

For most Ugandan land the choice between Robusta and Arabica coffee farming is already made, because the two species occupy different altitudes and barely overlap. Robusta takes the low ground and is about 80 percent of the national crop; Arabica takes the highlands and earns more per kilogramme. Only a narrow band of country makes this a genuine decision.

If you are in that band, the usual advice is not much help. It says Robusta yields more and Arabica pays more, which is true at one management level and stops being true at another. The measured Ugandan evidence points somewhere more interesting, and in one place it points the opposite way.

The Altitude Gate: Where Coffee Growers Actually Have a Choice

Start with the figure that removes the question for most readers. The two published bands are Robusta on low ground and Arabica in the highlands, and how much they overlap depends on which passage of the ministry's own handbooks you read.

The published bands, both readings, because the documents give each species two.
Robusta. Two passages of one handbook say 900 to 1,500 metres. One passage of the same handbook, and the clonal nursery manual, say 900 to 1,200.
Arabica. Two passages say 1,200 to 2,500 metres. Its species description says 1,300 to 2,500.
On the widest readings the two overlap between 1,200 and 1,500 metres, a band 300 metres deep. That is where the choice is real.
On the narrowest readings Robusta stops at 1,200 and Arabica starts at 1,300, so there is no overlap at all and a 100 metre gap where neither species is stated to fit.
What to do with that. The altitude figures are not settled, so do not lean on them alone. The temperature ceilings are stated once each and consistently, and they are the more reliable test.

Those ceilings do more work than the altitude figures. Set the two published optima side by side and they barely touch: 22 to 28 degrees for one species, 15 to 24 for the other. The overlap is two degrees. Past 28 Robusta still has room before its stated damage point at 30; past 24 Arabica has none at all, because 24 is where its handbook puts the limit. So the mid twenties is comfortable ground for one species and already beyond the other, and that is a sharper test than any altitude band because each figure is stated once rather than twice.

Soil runs the opposite way to intuition. The lowland species asks for more depth, 1 to 1.5 metres of working soil where it is moist and up to 3 metres where it is dry, while the highland species asks for roughly a metre. Neither tolerates waterlogging and both want the same slightly acidic pH, so depth is the one soil variable that separates them, and it favours Arabica.

Yield Per Acre: The Robusta Advantage Shrinks as Farming Improves

One Ugandan field study measured both species at three management levels on the same footing, which makes it the only source that can answer this properly. Clean coffee per acre per year came out like this: Robusta about 232 kg under traditional practice against Arabica's 180 kg; Robusta about 505 kg under improved practice against 473 kg; Robusta about 841 kg under recommended practice against 805 kg.

The same three pairs read as a percentage. This is the comparison that matters and nobody publishes it.
Traditional practice. Robusta yields about 29 percent more per acre than Arabica.
Improved practice. The gap falls to about 7 percent.
Recommended practice. The gap falls to about 4 percent, which is inside the noise of a measurement this size.
What that means. "Robusta is the higher yielding species" is a statement about badly managed gardens. Bring both species up to recommended practice and they produce almost the same weight of coffee off an acre.

The mechanism behind it is a pair of numbers pulling in opposite directions. Robusta is planted at roughly 450 trees an acre and Arabica at about 680, so Arabica puts about half again as many trees on the same ground. Divide the traditional figures by those populations and Robusta gives about half a kilogramme of clean coffee a tree against Arabica's quarter kilogramme. Robusta yields about twice as much per tree; Arabica plants about half again as many trees. That is why the per acre gap is so much narrower than the per tree gap, and it is the single most useful sentence in this comparison.

The Arabica per tree figure has an independent check behind it. Its handbook separately gives national average productivity as 0.3 kg of clean coffee per tree per year, from a completely different direction, and the study's traditional figure divided by 680 trees lands in the same place. Two unrelated sources agreeing is rare on this crop. Everything above the measured figures, including what the variety tables promise and how far above field reality those promises sit, belongs to the coffee yield per acre page; the two tree counts are derived from the spacing geometry on the coffee spacing page.

Price Against Cost: Arabica Earns More Per Acre and Keeps Less of It

The same study recorded what each group of farmers was paid and what they spent. No shilling figures appear on this page, for a reason given at the end of this section, but the ratios are the finding and they are clear.

Arabica's producer price ran about 39 to 52 percent above Robusta's across the three groups, with the premium at its widest among traditional farmers and its narrowest among those following recommended practice. Because of that premium, Arabica's gross income per acre came out ahead of Robusta's at all three levels, even where it produced less coffee.

Then the cost side moves, and it moves hard. Among traditional farmers Arabica's variable cost was about 39 percent lower than Robusta's, because those growers were using family labour and buying almost nothing. Among improved farmers it was about 43 percent higher. Among those following recommended practice it was about 87 percent higher, which is very nearly double, driven by hired labour, bought inputs and a planned programme against the two highland diseases.

Follow that through to what is left. Gross margin as a share of revenue was 88, 73 and 70 percent for Robusta across the three groups, and 94, 71 and 57 percent for Arabica. Arabica starts better and ends thirteen points worse. Its gross margin per acre stays ahead throughout, but the lead narrows from about 25 percent at traditional practice to about 30 percent at improved and then to about 9 percent at recommended, against a cost base nearly twice as large.

So the honest answer to "which is more profitable" is this. Arabica is paid more for less coffee, and the more a grower spends closing the yield gap, the less of that extra price they keep. A grower who can finance the cost side is choosing between two similar margins with very different amounts of cash at risk. A grower who cannot is better served by the species whose costs rise more slowly. Working out which of those two you are is a better use of an afternoon than comparing yields, and the break even method is the tool for it.

Why no shillings here. Those figures are one study's producer prices from a single period, they would read as current if printed without that period, and they would sit awkwardly beside the cost of coffee farming page, which keeps the budgets and deliberately publishes no current price for the same reason. The ratios survive any exchange rate and any season, which is what makes them worth printing. For current levels by product, the live coffee price page is the place to look.

How Much Weight the Measured Farm Figures Can Carry

Everything in the two sections above rests on one study, so its size belongs in the open rather than in a footnote.

What the study actually sampled, from its own sample frame.
96 farmers in total. Six sub regions, one district in each, one sub county in each district, four management categories, four farmers per category.
Four of the six districts were Robusta and two were Arabica. So each Robusta figure rests on roughly 16 farms and each Arabica figure on roughly 8.
The two Arabica districts make different products. One is a dry processed district and the other a washed district, so the Arabica column averages across two crops that are sold in different forms at different prices.
What follows. Trust the direction of travel, which is consistent across all three levels and matches what the study itself says in words. Do not build a budget on the size of any single gap, and treat the Arabica column as the weaker of the two.

That is why this page argues from percentages rather than handing over a verdict. The direction is large enough to act on: the yield gap narrows, the price premium narrows, the cost gap widens. The exact size of each is not.

Planting Material: A Ugandan Bred Clone Against a 1950s Selection

This comparison has nothing to do with money and it may matter more than anything above, because it decides what a grower can walk into a nursery and buy.

Question Robusta Arabica
Raised from Cuttings Seed
Breeding system stated Self sterile Not stated
Variety mixing rule Five lines or more None given
Bred in Uganda Ten released lines None in the table
Covered against both main diseases Not applicable Two of six entries

The difference is easiest to see as two shopping trips. A Robusta grower can ask for ten named lines bred at Uganda's own coffee research institute and released within the last two decades, every one of them carrying recorded resistance to the disease that nearly ended the crop. The catch is a rule that comes with them: because the species will not fertilise itself, the published instruction is five different clones or more in one production garden, so the trip is for a labelled mixture and not for a single favourite. How that works is set out on the Robusta growing guide.

An Arabica grower gets six entries, four of them mid century selections from a regional quarantine station and two credited to Kenyan breeders, and not one of the six is credited to Uganda in the column that records where each came from. Two of the four the same handbook's prose calls recommended and authorised carry remarks in its table refusing them for cultivation. Of the two entries clear of both highland diseases, one dates to 1970 and the other to 1985, and both are recorded as reaching only a handful of growers in one area. Row by row treatment is on the Arabica growing guide.

Two limits keep that comparison honest. The Arabica handbook says nothing about the species' breeding system and issues no mixing rule, and that is a gap in a document rather than a property of the plant, so read no conclusion into it either way. And a missing Ugandan credit in that table says what the published list contains, not what exists in every research record in the country.

The age of the Arabica list has a published explanation. Work reviewing coffee breeding programmes internationally identifies a development time of over twenty years per variety as one of the field's hardest operational constraints. On that clock a list assembled from mid century material is what you would expect to still be holding, and nothing a grower does this season shortens it.

Disease Pressure Is Not Symmetrical Between the Two Species

Both species have a serious disease problem. The difference is that one of them has been answered and the other has not.

Robusta's problem was coffee wilt disease, identified in 1993, which had destroyed about 56 percent of Uganda's Robusta trees by 2002 and took the six clones then recommended down with everything else. Breeding produced ten resistant lines and those are now the recommended material. It cost the country a decade and most of its trees, and it worked.

Arabica's problems are coffee leaf rust and coffee berry disease, both highland diseases, and the ministry's own variety table records the widely available older material as very or highly susceptible to one or both. No breeding answer has reached the published Ugandan list. The handbook is direct about the consequence in its own comparison of the two species: it states that Robusta is less susceptible to pests and disease and needs fewer fungicides and pesticides than Arabica. That is the ministry comparing its two crops in its own words, and it is the only such statement in either handbook.

One knock on effect reaches into the shade decision. Shade cuts heat stress and frost risk, and Arabica needs both, but the handbook also lists higher leaf rust risk among the drawbacks of shade trees. So shade on Arabica is a balance to be managed rather than a good to be maximised, where on Robusta the calculation is simpler. Ministry guidance on whether to spray at all splits along the same line, and the split is set out with its own thresholds on the coffee pest and disease control page. Nothing is named here: a product, a strength and a waiting period come off the container, checked with someone who has seen the garden.

Labour and Processing: Two Different Jobs on the Same Farm

The two species are trained differently from the first year. Robusta is bent along the row and pegged down so that suckers break from the base, ending on three stems. Arabica is nipped, meaning the tip is clipped off the young seedling to force stems from lower down. Different tools, different timing, different skill, and neither is a minor detail because a stem count cannot be changed later without losing a season.

At the other end of the year the split is larger. Most Ugandan Arabica is wet processed, and that is a capital question rather than a technique question: machinery, a large and clean water supply, drying space and a dedicated store, all bought before the first crop pays for anything. Robusta mostly leaves the farm as whole dried cherry instead, and so does Arabica in the one exception the handbook names, covering Kasese and the south west. So two Arabica growers in different districts face different capital bills for the same species, while a Robusta grower faces the lighter bill almost everywhere. Method detail is on the coffee processing methods page and picking standards on the harvesting and post harvest page.

Two things are the same and worth saying so. Both species carry two harvests a year, a main crop and a smaller second one, on a calendar that differs by region. And both are renewed by cutting back rather than replaced, staying productive for 40 to 50 years through repeated change of cycle, with the same three physical signs saying a bush is due. The pruning guide covers the operation for both.

What the Ministry's District Counts Do and Do Not Tell You

A reader wanting to know whether both species are realistic on their land naturally asks how many districts grow both. Three ministry documents answer, and they answer differently.

The Robusta handbook gives 60 districts growing only Robusta, 13 only Arabica and 38 both, out of a stated 104. Those three add to 111. The clonal nursery manual gives 66, 13 and 25 out of 104, which adds to 104 exactly. The Arabica handbook gives 44 Arabica districts, 28 growing only Arabica and 15 growing both, and 28 plus 15 is 43 rather than 44. Its percentages do not fit its own base either: it calls 44 districts 51 percent, which against 104 would be 42 percent, and solving each of its three percentages for a total gives something near 87.

So only the nursery manual's split adds up to its own stated total, and the figure these documents disagree about most is precisely the one this page is about. How many districts grow both species comes out as 38, 25 or 15 depending on which you open. Two of the three agree that 13 districts grow only Arabica, with the Arabica handbook the outlier at 28.

The useful conclusion is not a number. It is that a district list cannot settle this for you, and your own altitude and mean temperature can. Both handbooks agree on the two things that matter more than any count: Robusta is about 80 percent of national volume and Arabica about 20, and they occupy different ground.

Working Through the Choice for Your Own Garden

Four questions, in this order, because each one can end the decision before you reach the next.

Take these in sequence. Most growers stop at the first.
One: what is your altitude? Below about 1,200 metres, Robusta. Above about 1,500, Arabica, and that is also where the quality premium band starts. Between the two, keep going down this list.
Two: what is your mean temperature? If it sits in the mid twenties or above, Arabica is past the point where the handbook records photosynthesis ceasing, whatever your altitude suggests. This figure is stated once and consistently, so trust it over the altitude bands.
Three: what can you actually buy? For Robusta, ask whether five different certified clones are in stock near you, because the published instruction needs five and one line across an acre sets fruit badly. For Arabica, ask which of the six table entries is certified and available, and what its recorded disease susceptibility is. If the honest answer to either is nothing good, that settles it more firmly than any yield figure.
Four: can you carry the cost side? Arabica's advantage is a price advantage that only converts into a margin advantage if you spend, and at recommended practice its variable cost runs nearly double Robusta's for a margin lead under ten percent. If bought inputs and hired labour are not realistic for you, the species whose costs rise more slowly is the better bet even where Arabica would pay more per kilogramme.

Whichever way it goes, get the district agricultural office to confirm what is certified locally before buying, and a soil test before the first hole. Both species are perennials that stay in the ground for decades, so the establishment decisions are the ones that cannot be revisited. For the wider picture, the general coffee farming guide covers both species at overview depth and the export crops hub sets coffee beside Uganda's other export crops.

Questions Growers Ask When Choosing Between the Two

Which coffee is more profitable in Uganda, and what will each cost to establish? Published Ugandan data cannot settle the first part at the precision a grower needs, and the two halves of the question are connected, so take them together. The one field study measuring both species found Arabica's gross income per acre ahead at every management level, entirely on the strength of a producer price running about 39 to 52 percent above Robusta's, which comes mostly from slower ripening at altitude and partly from wet processing. But its variable cost at recommended practice ran about 87 percent above Robusta's, so its gross margin lead narrowed to about 9 percent and its margin as a share of revenue ended thirteen points behind. That study sampled about 16 Robusta farms and about 8 Arabica farms per management level, so trust the direction and not the size. On establishment, both species run several years before a real crop and the line items differ, with Arabica in a washing district adding equipment and a water supply that Robusta does not need. Itemised budgets are on the cost of coffee farming page and current levels by product on the live coffee price page; material prices sit on the seedling supplier page.
Can I grow both on one farm? Only where your land sits in the overlap, which on the widest published reading is 1,200 to 1,500 metres and on the narrowest does not exist. Ministry documents say between 15 and 38 districts grow both, which is too wide a spread to plan on. Growing both also means running two training systems, two disease programmes and possibly two processing routes, so it is more than twice the management rather than the same work spread over two crops.
Does Arabica really yield less than Robusta? Per tree, yes, by roughly half. Per acre the gap is much smaller because Arabica is planted at about 680 trees an acre against Robusta's roughly 450, and the gap narrows further as management improves: about 29 percent at traditional practice, about 7 percent at improved and about 4 percent at recommended.
Which is easier for a first time grower? Robusta, on the published evidence. Its handbook states that the species is less susceptible to pests and disease and needs fewer fungicides and pesticides than Arabica, resistant material bred in Uganda is available for its main disease, and its costs rise more slowly as management improves. The catch is the five clone rule, which is easy to get wrong and invisible when you do.
Does one of them need more work through the year? Arabica, on most farms. The two species are trained differently in the first year, Arabica carries the heavier disease programme because no widely available variety is covered against both of its main problems, and in a washing district its crop has to be pulped, washed and dried rather than simply dried whole. Both species carry two harvests a year, so both need labour found twice rather than once. If hired labour is hard to come by where you are, that asymmetry matters more than any yield figure.
Which one gets more out of a small plot? Arabica puts more trees on the same ground, about 680 an acre against roughly 450 for Robusta, and on a small plot that matters more than it sounds. It also means more planting material to buy, more holes to dig and more trees to prune and pick, and the measured per acre output of the two at good management is almost the same. So a small plot does not favour either species on weight of coffee; it favours whichever one you can keep on top of, since a traditionally managed acre of either produced roughly a quarter of what a well managed one did in the same study.
I already have one species. Should I replace it with the other? Rarely, and the yield argument does not support it: at good management the measured per acre gap between the two is only a few percent. Replacing means uprooting a perennial, losing years of production and starting establishment again. Set that against what the same study measured inside each species, where moving from traditional to recommended practice raised output by a factor of about three and a half on Robusta and about four and a half on Arabica. Improving the trees you already have is a far larger change than swapping which species is in the ground.

Before committing either way, confirm with the district agricultural office which certified material is genuinely available near you and check current prices by product, since a price premium you cannot reach and a variety you cannot buy both count for nothing. Knowing who buys which product in your district is worth settling in the same conversation, and a plant population count on your own layout will tell you how many trees you are really budgeting for.

More in Export Crops