How do I save for short-term goals without mixing the money?

The cleanest way to keep short-term goal money separate is to give each goal its own account and its own automatic transfer. When the money lives in the same pile as your checking, it gets spent by accident. When it lives in a labeled bucket with a monthly auto-deposit, it stops being a decision.

The mechanics are the same as a sinking fund for irregular expenses — but the vehicle choice matters more for short-term goals because the timeline is shorter. A vacation in six months, a wedding in 18 months, a down payment in 24 months — these need the money to be there, in cash, with no risk of loss right when you need it.

A practical decision tree by timeline:

0–12 months out (you’ll need it this year or next): Use a high-yield savings account (HYSA). Top rates as of mid-2026 are roughly 3.00%–4.15% APY depending on the bank, which is multiples of the national savings average of ~0.60%. The money is FDIC-insured (up to $250,000 per depositor, per bank, per ownership category) and accessible within 1–2 business days. Trade-off: returns don’t beat inflation, but you won’t lose principal right before you need it.

12–24 months out: A CD ladder can lock in slightly higher rates than an HYSA in exchange for giving up some liquidity. Vanguard, Fidelity, and most brokerages offer brokered CDs with rates up to ~4.35% APY on shorter terms (E*TRADE lists 4.35% on 6- and 9-month CDs as one recent example). The ladder structure — splitting the money across CDs maturing in 3, 6, 9, and 12 months — keeps some cash accessible at all times. The downside: breaking a CD early triggers an interest penalty, so this only works if you’re confident you won’t need the locked portion before it matures.

24+ months out, but still “short-term”: Treasury bills (T-bills) purchased through TreasuryDirect or a brokerage are competitive with brokered CDs, currently around 3.80%–4.02% on 1-year bills. They’re backed by the U.S. government, so there’s no credit risk, and you can sell them on the secondary market if you need the cash before maturity (the price will fluctuate with interest rates, so you might take a small loss — but you won’t lose principal at maturity).

What you should not use for money you’ll need within 24 months:

• Individual stocks or stock-heavy ETFs. A 24-month window can absolutely include a 20%+ drawdown — the S&P 500 dropped more than 30% from peak to trough multiple times in the last 25 years. If you’d be forced to sell during a downturn, you don’t have a savings goal; you have a forced-loss scenario.
• Crypto or crypto-adjacent assets. Same problem on a longer timeline; worse on a 24-month one.
• Long-term bonds or bond funds with duration >5 years. Interest-rate risk can erase a year of yield quickly.

A setup that works for most households:

1. Open one HYSA labeled “short-term goals” at a bank with no minimum balance and no monthly fee (Marcus, Ally, Wealthfront, AMEX, Discover, and most online banks qualify).
2. Automate a transfer from checking on each payday. Pick a number that doesn’t strain — $100/month per goal is a working baseline.
3. Resist the urge to consolidate goals into one bucket. Two HYSAs cost the same as one and the mental separation is the whole point.
4. When the goal arrives, transfer the money back to checking and pay the bill. Don’t close the account; you’ll need it for the next goal.

If you are juggling two or three short-term goals at once (say, an emergency fund, a vacation, and a car repair fund), the structure gets easier when you treat each as its own small project rather than one combined pile. The mental model: every goal has a target date, a target amount, and its own monthly contribution. When the contributions are automated, the only thing left to decide is whether you want to accelerate any of them — and that’s a decision you can make quarterly instead of weekly.

The complementary question is the vacation and irregular expenses answer, which walks through the same separation discipline for the specific case of expenses you can predict but not control — and where the right vehicle may be a CD ladder rather than a single HYSA.