Trampoline Park Parking: 90 to 375 Spaces for the Same Box
Prince George’s County, Maryland, required 355 parking spaces for a 28,328 sq ft (2,632 m²) trampoline park in a leased warehouse shell. The building already existed, and the only site work proposed was restriping the lot.
The operator’s consultant told the planning board the site needed 52 spaces at capacity, or 86 on a deliberately conservative reading. Staff recommended approving a departure of 216 spaces, which would leave a requirement of 139 against the 177 spaces the lot could be restriped to hold.
Put the same building in Hickory, North Carolina, and the number becomes 85. Nothing about the trampolines changes. The number on the site plan does.
Parking is the cheapest thing to overlook when you tour a box and the most expensive thing to fix after you sign. This article works through what four US codes actually require, what operators have reported observing in their own parks, and what the difference costs when it has to be built.
1. Four codes, four different answers
There is no national parking standard for a trampoline park in the United States. Each jurisdiction classifies the use and applies a ratio from its own schedule. The classification is a judgement call, and the ratio inherits that judgement.
Dallas is the clearest example of one code disagreeing with itself. A children’s amusement centre is parked at one space per 200 sq ft. Anything that falls into the residual “other uses” bucket inside commercial amusement is parked at one space per 100 sq ft. Same building, same city, double the stalls, decided by which line the plan reviewer ticks.
| Jurisdiction and source | Use classification | Ratio required | Sq ft per stall |
|---|---|---|---|
| Dallas, Texas (city zoning page, commercial amusement inside) | Children’s amusement centre | 1 per 200 sq ft | 200 |
| Dallas, Texas (same page) | Residual commercial amusement line | 1 per 100 sq ft | 100 |
| Prince George’s County, Maryland (section 27-568, as applied in staff report DPLS-448) | Other commercial recreation | 1 per 80 sq ft of patron area | 80 |
| Hickory, North Carolina (Land Development Code, Schedule A) | Amusement facilities, indoor | 3 per 1,000 sq ft | 333 |
| Plymouth, Minnesota (ordinance as summarised in a 2024 staff report) | Indoor commercial recreation | 1 per 300 sq ft of 90% of floor area | 270 effective |
| American Planning Association, PAS Report 510/511, cited in the Maryland file | Indoor amusement facility | 3 per 1,000 sq ft | 333 |
Two details in that table matter more than the spread itself.
The first is arithmetic. Plymouth’s rule is 90% of floor area divided by 300. Ninety percent of a number, divided by three hundred, is the same as three per thousand. Plymouth and Hickory and the national guidance all land on the same effective requirement without any of them saying so.
The second is how the ratios were chosen. In Dublin, Ohio, the applicant for a 24,000 sq ft Rockin’ Jump asked to use one space per 300 sq ft because the code had no trampoline category at all and the nearest analogue was a skating rink. Staff accepted it. The number came from a different business with a different visit pattern.
Hickory shows what the schedule thinks of the use. Indoor amusement sits at 3 per 1,000, level with general retail, below health clubs at 5 per 1,000 and well below restaurants at 8 per 1,000. A code that files your attraction in the same bracket as a shop is telling you something about how it models your peak.
2. Turn the ratio into stalls before you tour the site
The calculation is one division, and it belongs at the start of site selection rather than the end. Take the floor area the landlord will certify, divide by the sq ft per stall, and round up.
For a 30,000 sq ft (2,787 m²) box: 30,000 divided by 80 is 375. 30,000 divided by 100 is 300. 30,000 divided by 200 is 150. And 30,000 multiplied by 3 and divided by 1,000 is 90.
| Requirement | 20,000 sq ft | 30,000 sq ft | 40,000 sq ft |
|---|---|---|---|
| 1 per 80 sq ft | 250 stalls | 375 stalls | 500 stalls |
| 1 per 100 sq ft | 200 stalls | 300 stalls | 400 stalls |
| 1 per 200 sq ft | 100 stalls | 150 stalls | 200 stalls |
| 3 per 1,000 sq ft, which is Plymouth’s 90% rule | 60 stalls | 90 stalls | 120 stalls |
At 30,000 sq ft the top and bottom rows differ by a factor of 4.2. That is not a disagreement about traffic volumes. It is a difference in what each code thinks a trampoline park is.
The area you divide matters as much as the divisor. Prince George’s County applies its ratio to patron service area rather than gross floor area: the applicant allocated 28,328 sq ft of the 30,520 sq ft tenancy to patrons and the rest to back-of-house, which took the requirement from 382 stalls to 355. In Plymouth the ratio applies to 90% of floor area, a haircut written into the ordinance for storage and circulation.
So the useful number is not the ratio on its own. It is the ratio applied to the certified area your landlord will put in the lease. Ask for that figure in writing before the letter of intent, because it is the denominator in every parking argument you will have for the next ten years.
3. What operators report actually parking
Four projects have enough public paperwork to compare what the code asked for with what the site held or what the operator said it needed. None is a national sample, and two of the demand figures come from the applicant rather than the city. They are still the only measured numbers in the public record.
| Project and record | Floor area | Stalls required or estimated | Stalls available | Peak demand in the record |
|---|---|---|---|---|
| Launch Trampoline Park, Capitol Heights, Maryland (DPLS-448, September 2017) | 28,328 sq ft patron area | 355; staff recommended a departure to 139 | 177 | 52 vehicles at four patrons a car, 86 at two; 136 patrons and 18 staff at capacity |
| Rockin’ Jump, Dublin, Ohio (case 16-010CU, March 2016) | 24,000 sq ft | 100, being 80 at one per 300 plus 20 office stalls | 104 | 120 jumpers, 20 staff and 40 to 60 spectators at capacity |
| Indoor playground, Plymouth, Minnesota (planning report, April 2024) | 20,000 sq ft play area | 81 across the whole building | 60, plus 14 by restriping | 50 vehicles on a busy winter weekend, 30 on a typical winter weekday |
| Get Air, Turlock, California (as reported by the Turlock Journal) | 12,280 sq ft | No code figure in the record; staff estimate of 245 potential demand | 426 in a shared centre | Under 40% of the 426 occupied at the park’s peak |
Converted to a common unit, the observations cluster while the requirements do not.
Plymouth’s 50 vehicles against 20,000 sq ft is 2.5 vehicles per 1,000 sq ft, or one stall per 400 sq ft. The conservative Maryland figure of 86 against 28,328 sq ft is 3.0 per 1,000. Dublin provided 104 stalls for 24,000 sq ft, which is 4.3 per 1,000, and housed 180 to 200 people at capacity in them. Against those three numbers, codes asking for 10 or 12.5 per 1,000 are asking for three to four times what the operators were seeing.
The Turlock file is the outlier in the other direction, and it is instructive. City staff put the park’s potential demand at 245 stalls for a 12,280 sq ft suite, which works out to 20 per 1,000, five times the highest ratio in any code in this article. On that figure the park would have needed more than half the centre’s 426 spaces. The applicant then submitted a parking study showing the centre’s own peak at noon on Thursday and Saturday and on Tuesday between 5:30 and 6 p.m., against a park peak of 6 to 9 p.m., and the newspaper reported that even at peak, occupancy would stay below 40% of the lot.
That is the whole argument in one case. The staff number was a ratio applied to floor area. The applicant’s number was a clock.
4. Why the code number and the observed number diverge
Four mechanisms explain most of the gap, and all four are visible in the records above.
4.1 Cars carry families, not shoppers
A trampoline park is a group destination with a natural minimum of about two: one adult and one child. The Maryland operator argued four patrons per vehicle on the grounds that the park is reservation-driven and groups arrive together. The conservative reading used two. Plymouth’s observed 50 vehicles against a 20,000 sq ft playground implies roughly two people per car if the park holds about 100 guests.
Retail ratios were not built on that assumption. A code calibrated on shoppers arriving alone will over-provide stalls for a use where almost nobody arrives alone.
4.2 Half the floor is not occupiable
The Maryland filing put 14,200 sq ft of a 28,328 sq ft patron area into activity space, and the rest of that patron area still counted toward the parking ratio. At 136 patrons in the activity areas, that is about 104 sq ft of jumping surface per person at capacity. A ratio applied to gross floor area divides by floor that nobody is standing on.
4.3 The ticket sets the dwell time
A 90-minute session is a fixed two-hour commitment with a queue at the front desk. During that window one stall serves one car and does not turn. A shopping trip does not work that way: the car is in and out inside the hour, so a retail stall turns more than once in the same period. The ratio measures floor area, which is a proxy for people, and then forgets that the people stay.
4.4 Peaks do not coincide, and that is the whole opportunity
Turlock’s centre peaks at lunch and on Saturday midday. The park peaks from 6 to 9 p.m. A supermarket and a gym can share a lot with a trampoline park because their peaks are offset by hours rather than by luck. That is what shared parking means in practice: a schedule, not a favour.
The correlation trap: sites with generous parking are often the profitable ones, because both follow a strong trade area. That does not mean the stalls caused the revenue. Nothing in the public record links stall count to park revenue, and this article does not claim it does.
5. What the shortfall costs if it has to be built
Surface stalls are cheap and land-hungry. Structured stalls are the reverse. The moment a site cannot reach its ratio with asphalt, the marginal stall is priced like a building.
The most recent public numbers come from a Walker Consultants feasibility study for downtown Napa, California, reported by the Press Democrat in August 2026. The preferred site yields 390 to 460 spaces at $73,500 per space, about $34.2 million. The three alternatives came in at $95,000, $98,300 and $143,400 per space.
Apply the low figure to a 30,000 sq ft park and the requirement becomes a capital project. Each row below multiplies spaces required by $73,500, amortises the result over 30 years at 6%, and divides by 60,000 admissions a year, which is two visits per sq ft of floor area.
| Requirement at 30,000 sq ft | Stalls | Structure cost at $73,500 a stall | Annual debt service | Cost per admission |
|---|---|---|---|---|
| 1 per 80 sq ft | 375 | $27.6 million | $2.0 million | $33.37 |
| 1 per 100 sq ft | 300 | $22.1 million | $1.6 million | $26.70 |
| 1 per 200 sq ft | 150 | $11.0 million | $801,000 | $13.35 |
| 3 per 1,000 sq ft | 90 | $6.6 million | $481,000 | $8.01 |
Read the top row against a $19.99 ticket. Debt service alone is $33.37 an admission, which is 67% more than the ticket price. Even the bottom row, the most permissive requirement in this article, adds $8.01 an admission, or 40% of the ticket, before a single trampoline is bought.
The gap between required and observed is where the money sits. At 30,000 sq ft the Maryland ratio asks for 375 stalls; the conservative operator estimate for a similar floor was 3 per 1,000, which is 90. The 289-stall difference costs $21.2 million to build and $1.5 million a year to carry, or $25.72 per admission on the same 60,000-visit base.
Land does not escape the arithmetic either. A surface stall, with its share of aisles, drive lanes and perimeter landscaping, takes roughly 300 to 350 sq ft of site. That is a planning convention rather than a code figure, and we flag it as such. At 325 sq ft, 90 stalls need about 29,250 sq ft, roughly the building’s own footprint. At 375 stalls they need about 121,875 sq ft (11,323 m²), four times the building. That does not fit on a padsite in front of a leased box.
This is why the parking number, rather than the rent, ends more early-stage deals than any other line. Our analysis of mall anchor conversions found the same thing from the landlord side: the box is available, and the lot is the constraint.
6. What the data does and does not show
Four mechanisms explain a gap. Four codes and four project records do not prove a national rule, and it is worth being precise about the limits.
The sample is thin. Three US codes, one national guidance figure and three project records, from Maryland, Ohio and Minnesota, are not a national survey. Two of the four demand figures, at Launch and at Plymouth, were produced by the applicant. An operator asking for a parking reduction has a clear incentive to describe low demand, which is why both numbers were tested against capacity: 136 patrons plus 18 staff at Launch, 120 jumpers plus staff and spectators at Dublin.
The follow-up is missing. A Prince George’s County staff report recommended a 216-stall reduction in 2017. Nobody has published what that lot actually held on a Saturday night afterwards. The one thing that would settle the question is post-opening occupancy counts from a dozen parks, and they are not in the public record.
The cost geography is coastal. The per-stall costs above come from California, where structured parking sits at the expensive end of the US range. A surface stall in a cheap land market costs a fraction of that. We have no public per-stall surface cost for a trampoline park, and we are not going to invent one.
The process is discretionary. A departure or variance is a discretionary decision. The Maryland staff report accepted the argument; another reviewer with another traffic engineer could have refused the same file. Reading one recommendation as a precedent is the most common mistake made with these records.
The editions move. Ratios change with code amendments. The figures here come from the editions listed in Sources. Before you rely on any of them, pull the current schedule for your address.
Two conclusions do survive. The observed requirement across three parks falls in a narrow band, 2.5 to 4.3 stalls per 1,000 sq ft, while code requirements span 3 to 12.5. And the cost of closing a large gap is big enough to change a deal, not to trim a line item.
7. Five questions to answer before the letter of intent
The order matters. Each question can end the process, and each is cheaper to answer than the one after it.
1. What does the code call this use, and which line does it pin? Dallas is the warning: children’s amusement centre is 1 per 200, the residual commercial amusement line is 1 per 100, and the difference on a 30,000 sq ft box is 150 stalls. Ask the planner to confirm the classification in writing before you commission drawings.
2. How many stalls does that ratio produce at the certified area? Divide, do not estimate: area divided by sq ft per stall, or area multiplied by stalls per 1,000 and then divided by 1,000. Check whether the ratio applies to gross floor area, patron area or a haircut of floor area, because that factor moved the Maryland requirement by 27 stalls.
3. What does the operator’s own experience say? Use 3 per 1,000 sq ft as a working upper bound for a reservation-driven park with parties, then test it against your capacity: guests at peak divided by people per car, plus staff. If your plan assumes four guests per vehicle and your party mix does not support it, you are borrowing cars that will not arrive.
4. When does the rest of the centre peak? If the other tenants peak at lunch and on Saturday morning, and you peak on weekday evenings and Saturday afternoon, the shared lot is doing real work. If you are the evening anchor in a cluster of restaurants and a cinema, your peak is the centre’s peak and the arithmetic gets worse, not better.
5. Who pays for the shortfall? Two paths exist. A departure or variance costs months of hearings and returns a discretionary answer, which is why it belongs in the lease as a condition rather than in the business plan as an assumption. Structured parking costs $73,500 to $143,400 a stall in the 2026 study cited above, and no single trampoline park should carry that: at the Maryland ratio the debt service alone runs to $33.37 an admission, which is more than the ticket price.
For the sizing side of the same decision, our layout breakdown for a 22,000 sq ft park shows how floor area splits between zones, and the five-filing cost comparison shows what the build-out itself runs per sq ft. Parking sits outside both, and it decides which building you can use.
8. Summary, and three moves before you sign
The requirement is a ratio applied to a floor area. Both inputs are negotiable in ways the resulting number is not. Codes ask for anywhere from 90 to 375 stalls on the same 30,000 sq ft box. Operators report 2.5 to 4.3 stalls per 1,000 sq ft. The gap is not a rounding error; it is a capital project.
1. Get the classification and the ratio in writing. One email to the planning counter, before the letter of intent, tells you whether you are solving a 90-stall problem or a 375-stall problem.
2. Count the lot at 6 p.m. on a Saturday. Count occupied stalls, not the stall count on the site plan, and count them in the hours you will actually trade. That single hour of observation is worth more than any ratio in this article.
3. Put parking in the lease, not in the hope. If a departure is needed, make the lease conditional on it. If the site needs a structure, make sure someone else is paying. The carry cost of the loan and the rent per sq ft are already fixed; parking is the variable you can still move.
If you are weighing two sites and the parking numbers differ by a factor of four, send us both addresses. We will tell you which layout fits the box you can actually lease, and what the build-out looks like at that footprint. Start with the design and delivery process.
FAQ
How many parking spaces does a trampoline park need?
It depends on the jurisdiction, and the range is wide. For a 30,000 sq ft park, the four codes cited here produce between 90 and 375 stalls: 3 per 1,000 sq ft at the low end, one per 80 sq ft at the high end. Measured demand in three US parks worked out to 2.5 to 4.3 stalls per 1,000 sq ft. Ask your planner which classification applies before you budget.
Can a trampoline park share parking with a shopping centre?
Yes, when the peaks do not collide. In Turlock, California, the centre peaked at noon on Thursday and Saturday while the park traded hardest between 6 and 9 p.m., and reported occupancy stayed under 40% of 426 spaces. Sharing fails when your peak is the centre’s peak, which is usually the case in a restaurant and cinema cluster.
What happens if a site does not meet the parking minimum?
You apply for a departure, variance or conditional use, and the outcome is discretionary. A Prince George’s County staff report recommended cutting a 355-stall requirement to 139 in 2017 after the operator documented 52 to 86 vehicles of real demand. Plymouth, Minnesota’s planning staff accepted 60 stalls against an 81-stall requirement, with a condition allowing 14 more by restriping. Both went through a discretionary review, so the approval belongs in the lease as a condition.
Does more parking make a trampoline park more profitable?
No public dataset shows that. Sites with ample parking and sites with strong revenue tend to be the same sites, because both follow a strong trade area. Treat a generous lot as a site-quality signal rather than a revenue lever. The measurable cost of over-providing is easier to see: at $73,500 a structured stall, 150 extra spaces is $11 million before interest.
How much does a parking structure cost per space in 2026?
A Walker Consultants study for downtown Napa, California, put the preferred option at $73,500 per space for 390 to 460 spaces, and the three alternatives at $95,000, $98,300 and $143,400 per space. Those are coastal California figures for public structures. Surface stalls cost far less, which is the reason site selection beats value engineering.
Sources
- Maryland-National Capital Park and Planning Commission, Prince George’s County Planning Department, “Departure from Parking and Loading DPLS-448, Steeplechase 95, Building D”, staff report dated 8 September 2017. pgatlas.com
- City of Plymouth, Minnesota, Planning Commission staff report on a conditional use permit for indoor commercial recreation, 3 April 2024. plymouthmn.gov agenda
- City of Dublin, Ohio, Planning and Zoning Commission, Case 16-010CU, Shier Rings Road, Rockin’ Jump Dublin, staff report dated 24 March 2016. dublinohiousa.gov
- City of Dallas, Development Services, zoning page for Commercial Amusement (Inside), required off-street parking schedule. dallascityhall.com
- City of Hickory, North Carolina, Land Development Code, Chapter 9, Section 9.2.2 Schedule A, current posted edition. hickorync.gov
- The Press Democrat, Edward Booth, “Study favors former Napa pump station for new parking garage”, 17 August 2026, reporting the Walker Consultants feasibility study. pressdemocrat.com
- Turlock Journal, “Planning Commission to consider trampoline park project”, undated archive copy of the pre-hearing report on the Get Air conditional use permit at 1350 W. Main St. turlockjournal.com