Ask ten Uber drivers in Johannesburg what Uber takes from their fares and you will get ten different answers — 25%, 30%, “it depends”, or simply “too much”. The uncomfortable truth: they are all right. Uber’s cut in South Africa is no longer one clean number, and the gap between what riders pay and what drivers actually receive has become the single biggest hidden cost in e-hailing.
This guide breaks down every layer Uber takes, runs a real R33,000/month example through the numbers, and shows you how to check your own effective commission rate — because the platform certainly will not do it for you.
Uber takes a per-trip service fee that in practice ranges from roughly 25% to over 40% depending on the trip, plus a separate booking fee charged to the rider that never reaches the driver. There is no single flat “commission rate” any more — which is exactly why drivers’ estimates vary so wildly.
When a rider pays for a trip, the money is split before it ever reaches your account:
| Layer | Who Pays It | What It Is |
|---|---|---|
| Service fee | Deducted from the fare | Uber’s percentage cut of each trip. Varies by trip type, distance and promotions. |
| Booking fee | Paid by the rider | A flat per-trip charge covering safety features and regulatory costs. Never appears in driver earnings. |
| Marketplace/other adjustments | Varies | Promotions, ride-type differences and regional variations that shift the effective rate. |
The booking fee is the detail most drivers miss. Because riders pay it on top of the fare, and it is stripped out before your share is calculated, a driver comparing “what the passenger paid” against their payout will always find a bigger gap than the headline service fee suggests.
South African drivers report effective totals — service fee plus all deductions — anywhere between 25% and 42% of the rider’s total payment. On a R33,000/month gross, the difference between 25% and 42% is R5,610 every month, or R67,320 a year.
Uber South Africa no longer publishes one flat commission figure, and drivers’ own trip data confirms why: the effective rate moves trip by trip. Here is what different effective rates mean for a driver grossing R33,000/month — the typical full-time Johannesburg driver working 55+ hours a week:
| Effective Commission | Platform Takes (per month) | Driver Keeps (per month) | Difference vs 25% |
|---|---|---|---|
| 25% | R8,250 | R24,750 | — |
| 30% | R9,900 | R23,100 | −R1,650 |
| 35% | R11,550 | R21,450 | −R3,300 |
| 42% | R13,860 | R19,140 | −R5,610 |
The annual version is worse: over a year of full-time driving, the gap between a 25% and a 42% effective rate is R67,320 — enough to buy a used Suzuki Dzire cash. Two drivers working identical hours on the same platform can effectively earn different wages because of how the fees land on their trip mix.
Uber is not alone in this structure — Bolt’s commission works differently in ways that matter, and drivers running both apps often find the effective rates converge more than the marketing suggests. The point is not that one platform is “the good one”. It is that the number on the advert is not the number on your payout screen.
You can audit this yourself in one evening:
Run it over four weeks, not one — a single surge-heavy or promo-heavy week will distort the picture. Most drivers who run this exercise for the first time discover their effective rate is 5–10 points above what they assumed. That discovery changes how you plan: at 40% effective commission you need to gross ~R55,000 to keep what a 25%-rate driver keeps from R33,000.
Commission is only one of four big levers in your take-home pay — the others are hours, fuel efficiency and vehicle cost. But it is the lever you control least, which is exactly why you must measure it precisely and optimise the levers you do control.
A full-time driver’s monthly flow at R33,000 gross and a 30% effective rate looks like this: R9,900 to the platform, roughly R10,300 in fuel at current inland prices (R29.88/L for Petrol 93), leaving about R12,800 before insurance, data, maintenance and car costs. If you are renting at R2,500/week, another R10,750 leaves — and the month ends near zero. This is why two drivers in identical cars, working identical hours, can end the month R6,000 apart: trip mix, ride types accepted, and the effective rate each ends up paying.
For the full picture of where every rand goes, see our complete expense breakdown for Uber and Bolt drivers, and our analysis of what drivers actually earn per hour once commission and costs are counted.
Ultimately the platforms will always take their share — that is the deal. The drivers who win are the ones who know their exact effective rate, work the peaks, and keep their costs brutal-low. Start by running your real numbers in the FleetCalc profitability calculator, which uses current South African fuel prices and lets you model different commission scenarios side by side.
🧮 Model My Commission Scenarios →Two factors quietly shift the rate from trip to trip. First, ride type: Go, UberX, Comfort and Black carry different fare structures and fee treatments, so a driver’s mix across a month changes the blended rate even if no single fee changes. Second, promotions and quest bonuses: earnings boosters are paid outside the standard fare split, which can drag your effective rate down (in a good way) during promo periods and leave the base rate exposed when they expire.
This is why the audit matters weekly, not annually. A driver who chases surge all month may see an effective rate of 27%, while the same driver doing long low-fare airport waits without surge sees 38% on identical hours. The platform did not change — the trip mix did.
| Weekly Gross | Kept @ 25% | Kept @ 30% | Kept @ 35% | Kept @ 42% |
|---|---|---|---|---|
| R5,000 | R3,750 | R3,500 | R3,250 | R2,900 |
| R7,700 | R5,775 | R5,390 | R5,005 | R4,466 |
| R10,000 | R7,500 | R7,000 | R6,500 | R5,800 |
Run this pattern against your own statements. A driver tracking monthly effective rates through 2026 recorded this sequence: 28.1% in January, 28.6% in March, 29.4% in May, 30.2% in July, 31.1% in September. No announcement, no change to “the commission” — just a slow accumulation of booking-fee adjustments, promo expiries and fare-structure tweaks. On a R33,000/month gross, that three-point creep is a R990/month pay cut that never once appeared as a negative line item.
Drivers who tracked it noticed by June and adjusted: shifting hours toward surge windows, pushing more trip volume into promo periods, and re-weighting days across platforms where the effective rate diverged. Drivers who never measured first learned about it from their bank balance.
Uber’s cut is no longer a single flat percentage. It takes a service fee per trip that in practice has ranged from around 25% to well over 40% of what the rider pays depending on trip type, promotions and fare structure, plus a separate booking fee charged to riders that never reaches the driver. The only reliable number is the one in your own monthly statements.
No. The booking fee is deducted before the driver’s share is calculated. It is charged to the rider to cover safety features and regulatory costs, and it never appears in driver earnings — which is why comparing rider payments to payouts always shows a bigger gap than the headline service fee.
Bolt has historically advertised a lower headline commission (around 20–25%), but like Uber it adjusts effective fees through service charges and promotions, so the real take varies trip by trip. Compare your actual statements in both apps over a full month before deciding where to drive.
You cannot negotiate the commission, but you can maximise what reaches your pocket: chase surge and peak-hour trips, protect the reward-tier benefits that reduce effective fees, audit your statements monthly for booking-fee creep, and drive the most fuel-efficient car you can so the kept share goes further.