Last August I listed my 2019 Subaru Outback on Turo, the peer-to-peer car-rental platform. The thinking was simple: the car sat unused four out of seven days, the insurance and registration were paid for either way, and Turo's marketing copy estimated a "potential $480 a month" for cars of similar make and age. I committed to six months of running the experiment before judging it. Six months in, the answer was clearer than I expected: it was not worth it for me, in this city, with this car. The pieces of the math that broke might still work for someone else.
Setup and the listing
The Subaru: 2019, 47,000 miles, clean, no aftermarket modifications. Purchase price four years prior was $24,000; market value in August was about $14,800. I listed it at $54/day weekday and $72/day weekend, which positioned it at roughly the 60th percentile of comparable cars in my zip code.
Turo offers several insurance tiers. I chose "Standard," which keeps the host (me) responsible for the first $250 of any damage and takes 25% of trip earnings. The "Premium" tier reduces host liability to $0 but takes 40%. I chose Standard, reasoning that the chance of a damage event in any given month was small. This was the first place my reasoning was wrong.
Gross revenue, by month
| Month | Trips | Days rented | Gross revenue |
|---|---|---|---|
| August | 4 | 13 | $695 |
| September | 3 | 9 | $485 |
| October | 2 | 6 | $324 |
| November | 1 | 4 | $216 |
| December | 0 | 0 | $0 |
| January | 1 | 2 | $120 |
| Total | 11 | 34 | $1,840 |
Gross revenue across six months: $1,840. Turo's 25% Standard-tier fee: $460. Net to me before any other costs: $1,380.
The "potential $480 a month" estimate would have required an average occupancy of about 11 days per month at my pricing. I averaged 5.7 days. The estimate, like the per-hour estimates on most side-hustle platforms, was based on a high-utilization scenario most listings do not actually achieve.
Costs Turo did not warn me about
Three categories of cost showed up that the listing flow had not foregrounded.
Damages. Two trips out of eleven returned the car with damage. The first was a parking-lot scrape on the rear bumper, $620 to repair, of which I paid the $250 deductible and Turo's insurance covered the rest. The second was a chipped windshield, which I paid $185 to replace because the cost was below the deductible and not worth the claim hassle. Total damage cost to me, out of pocket: $435.
Increased wear. Across the 34 rental days, the car covered approximately 4,200 extra miles. At a conservative cost-per-mile of $0.18 (depreciation, tires, oil, brakes), that is about $760 of additional wear that does not show up on any single bill but reduces the car's eventual resale value.
Detailing. I had the car professionally detailed three times across the six months. $130 per detail, $390 total. Some of these were because of specific renter messes; some were preventive.
Net of these costs:
| Gross revenue | $1,840 |
| Turo fee (25%) | −$460 |
| Damage deductibles | −$435 |
| Detailing | −$390 |
| Extra wear (estimated) | −$760 |
| Net after direct costs | −$205 |
I lost approximately $205 over six months. Possibly worse, depending on how the wear estimate plays out at resale.
The non-financial cost
The number above does not capture the thing that actually drove me to delist. Every rental day was a small, low-grade source of stress. I checked the app twice a day. I worried about the renters I had never met. I rearranged my own driving schedule around the booking calendar twice. After the first damage incident I spent four hours on Turo's claim flow, which was reasonable but exhausting.
The stress was not the renters; most were perfectly nice. The stress was the cognitive overhead of having a $14,000 asset out of my sight, in someone else's hands, on a marketplace I did not fully trust.
A side hustle that asks you to entrust a major asset to strangers needs to pay you for the worry, not just the use. Mine did not.
Would I do it again?
For this car, in this city, no. The math was marginal on paper and negative when I included realistic wear. The stress made the marginal case worse.
It might work for: a second car you genuinely do not need (not a primary vehicle); a car located in a high-demand area (an airport-adjacent zip code, where listings get 12+ rental days a month); a vehicle type that commands a premium (a pickup truck during the moving season, a convertible in a beach town). For a daily-driver Outback in a mid-sized northeastern city, the marketplace's pricing did not support the costs.
If you are about to list, do this first: total your fixed costs (insurance, parking), your variable per-day costs (wear, detailing, deductible-risk), and then look at the listing prices for similar cars in your zip code. If the realistic gross at six rental days a month does not clear those costs by at least 40%, the experiment is not worth the worry.