The border is running 27% worse than last year
By Alan Scott · Who makes this · Tijuana, Baja California ·
If it feels like the line has gotten worse, it isn't in your head.
The comparison
Comparing the same months, January through July, in the General car lane at San Ysidro northbound:
| Median Jan–Jul | |
|---|---|
| 2025 | 74 min |
| 2026 | 93 min |
| Difference | +20 min · +27% |
Twenty minutes more, every single crossing. If you cross five days a week, that's an hour and a half of your life per week that didn't exist last year.
And month against the same month, it's worse than it looks
🔴 Each month is compared to the same month of 2025, not to the month before it. Comparing July to March of the same year measures summer, not change — the border always slows going into season.
| Month | 2025 | 2026 | Change |
|---|---|---|---|
| January | 67 | 59 | −12% |
| February | 55 | 80 | +46% |
| March (1–26) | 65 | 71 | +10% |
| April | ⚠️ | ⚠️ | the source published nothing |
| May | 80 | 96 | +20% |
| June | 78 | 121 | +56% |
| July | 89 | 120 | +34% |
| August (1–9) | 81 | 131 | +62% |
January was better than last year. From May on, every month worse — and June was the hardest hit of the year.
⚠️ April is absent because it doesn't exist in the data — not for us and not for anyone: the open data contains no records at all from March 27 to April 15, 2026. We don't fill it in. Here's the detail
The twist we didn't expect
Across four years, the story isn't "it gets worse every year":
| January–July | Median |
|---|---|
| 2023 | 92 min |
| 2024 | 85 min |
| 2025 | 74 min ← best of the four |
| 2026 | 93 min ← worst |
It's not that the border has been deteriorating year over year. It had been improving. 2025 was the best year on record. Then 2026 erased it completely: back to 2023 levels.
Which makes the question more interesting: something changed this year. This isn't slow, cumulative decay.
🔴 And San Ysidro is the LEAST affected of the three
When we first wrote this we had only measured San Ysidro. Measuring all three showed San Ysidro got off lightest:
| Crossing | Jan–Jul 2025 | Jan–Jul 2026 | Change |
|---|---|---|---|
| San Ysidro | 74 min | 93 min | +27% |
| Otay Mesa | 56 min | 75 min | +34% |
| Tecate | 44 min | 80 min | +84% |
Tecate doubled. And there's a detail that rules out the easy explanation: at Otay, the freight lanes got faster while the car lanes got 34% worse.
The full measurement of all three is here
What to do with this
Not much, honestly. You can't negotiate with the border.
But it's good for two things. One: if you're budgeting your time based on last year, you're 20 minutes short. And two: it means you can stop feeling like it's your bad luck. It isn't.
What this data does NOT tell you
It's San Ysidro's General car lane only. The other lanes moved far less — the Pedestrian Ready Lane didn't move at all.
It's a full-day average, so it doesn't tell you whether the extra 20 minutes landed in the morning commute or spread across the day.
🔴 And we don't know the cause. We have the what, not the why. We won't speculate about staffing or policy with data that can't support it.
🔴 Correction, August 12, 2026
The first version of this article said 25% and +18 minutes. It was wrong for two reasons:
- It mixed averages with medians — the 94 was an average and the 76 a median, so the percentage wasn't comparing like with like.
- It didn't exclude the 20 days the source never published, in either year.
Redone with medians in both years and the gap excluded from both: +20 minutes, +27%. It came out worse than we said, not better. It's on our public list of mistakes
You might also want:
- Something changed in 2026 and only the car lanes broke
- Friday is the best day to cross
- Why the Tuesday after a U.S. holiday is the worst day
Sources
Data: Average Daily Border Waiting Time, SANDAG Open Data Portal, from U.S. Customs and Border Protection records. Filter: San Ysidro · Passenger_vehicle_Standard_Lane · northbound. Period: January 29, 2023 to August 9, 2026. Method: medians, never averages. Same-month comparison between years. 🔴 The 20 days from March 27 to April 15 are excluded from every year — not just 2026 — because the source published nothing for that stretch in 2026. March is trimmed to the 26th in both years. Retrieved: August 12, 2026.
We are not affiliated with CBP or SANDAG.
CHARTS · build spec
Same two charts as the Spanish version — see 6_CONTENIDO/articulos/03_la_garita_va_peor.md. 🔴 Do not re-derive the values.
Label translations only, plus the rules that stay: 🔴 April is a visible gap labeled "the source published nothing" — not interpolated, not skipped. 🔴 August keeps reduced opacity and the asterisk, footnote "August is 9 days against the same 9 days."