If you live anywhere between Buhweju and Lake Victoria, this river is probably the reason there is water in your tap, grass on your grazing land, or a crop in your field. Most people who depend on it have never seen the whole thing.
Where it starts and where it goes
The Rwizi rises in the hills of Buhweju district, in the highlands of south-western Uganda. It is not a single spring but a gathering of small streams off steep, once-forested slopes. Those streams meet, and by the time the river reaches the flatter ground east of Bushenyi it is recognisably a river.
From there it runs broadly east and south-east, past Sheema and through Mbarara City, which draws its drinking water from it. Below Mbarara the gradient flattens further and the channel begins to meander through wide seasonal wetlands. It feeds Lake Mburo, then Lake Kachera, then Lake Kijanebalola, and finally reaches Lake Victoria through the wetland systems around Rakai and Kyotera.
Along that course it collects tributaries off the Nkore hills — streams out of Ntungamo and Sheema — and drains a network of papyrus swamps whose names rarely appear on maps: Nyakafumura, Kanyabukanja, the Mushasha basin. Those swamps are not incidental to the river. They are the part of it that stores water in the wet season and releases it in the dry one.
- Core catchment
- 2,282 km²Indicative: Inside a wider basin usually given as around 8,000 km².
- Mean discharge
- 25.4 m³/sCited: Measured at Mbarara. Range 16.5 to 35.7.
- Districts
- 10Cited: From Buhweju in the west to Kyotera in the east.
- Elevation at mouth
- 1,258 mCited: Where it enters the Lake Victoria system.
- Measured: Measuredcounted by us, in the field
- Cited: Citedpublished elsewhere, cited here
- Indicative: Indicativean approximation, not a measurement
Two catchment figures, both correct
You will see the Rwizi catchment given as 2,282 km² and also as roughly 8,000 km². The smaller figure is the river's own drainage area. The larger one is the water management zone that includes the connected lakes and wetlands downstream. We use whichever is appropriate and say which.
What the river actually does
Mbarara City is the largest urban centre in western Uganda and the Rwizi is its water supply. Not one of several — the supply. Add the towns and trading centres along the course, the institutions that abstract directly, and the industrial users, and you have a single river carrying an unusual amount of responsibility for its size.
The industrial and institutional list is a matter of public record: a brewery, soft drink bottling, dairies, hospitals, hotels, abattoirs, a research institution, universities and schools. Some of these abstract water. Most of them also discharge.
Beyond the piped supply, the river waters livestock across some of Uganda's most important rangeland, irrigates smallholder plots, supports fish farming, and feeds the wetlands that in turn support Lake Mburo National Park and two Ramsar-listed wetland systems — Lake Mburo–Nakivali and the Sango Bay forests.
The dry season is the test
Average flow is a comforting number and a slightly misleading one. What matters for a water supply is the minimum, and at Mbarara the recorded minimum is 16.5 m³/s against a mean of 25.4.
That gap is where the pressure lives. In a dry year, abstraction that is comfortable in April becomes contested in July. Local reporting has described the river dropping around two and a half metres below its historic level, and residents in some stretches describing a river they can walk across where they used to swim.
The wetlands are what buffer this. Papyrus swamp holds water back through the wet season and lets it go slowly. Drain the swamp for cultivation and you get the opposite: faster floods, then a harder dry season.
The four things worth understanding
It is a small river doing a big job
A 2,282 km² catchment supporting a city, an industrial base and several million rural people is a thin margin. There is not much slack in the system.
The wetlands are the storage
There is no large reservoir on the Rwizi. Papyrus swamp does that job. Every hectare converted to cultivation removes storage capacity that nothing replaces.
Damage is concentrated at the edge
The channel itself is fairly robust. What fails is the bank: cultivated to the waterline, grazed bare at watering points, or cut away by sand extraction.
Nearly all of it is reversible
Bank vegetation recovers if it is left alone and helped along. That is the whole reason this organisation exists rather than simply documenting a decline.
How the river changed
- 1960s
A forested catchment
Buhweju's headwaters carried substantially more forest cover. Flow was steadier and the dry season less severe. Population across the basin was a fraction of today's.
- 1980s–90s
Clearing and settlement
Land pressure pushed cultivation up the hillsides and out into the wetlands. Bank vegetation started coming out systematically rather than in patches.
- 2000s
Urban growth at Mbarara
Abstraction rose sharply while discharge into the river rose with it. Sand extraction became commercially organised rather than incidental.
- 2020
Mbarara becomes a city
City status brought investment, population and a further step change in both demand and effluent. The river is now the subject of national attention.
- 2020s
Restoration attempts
Government, NGOs and district authorities begin coordinated catchment work: bamboo in buffer zones, wetland restoration, mapped and demarcated boundaries. Progress is real and partial.
The detail is where this gets useful
The threats page sets out what is actually causing the damage, with sources. The water quality page has our monitoring readings, including the ones that got worse.