How To Make A Starter For Liquid Yeast

I’ve got a 2-liter flask that’s seen 40 batches and a stir plate that hums like a fridge with a sinus infection. A starter for liquid yeast is the difference between a beer that finishes at 1.010 and one that stalls at 1.024 and tastes like unfermented wort. This guide covers the three ways to build one, the gear that survives repeated use, and the one method I’d skip even if you’re broke.

Why you should trust me

I’ve brewed 50+ logged batches since 2014, three years as assistant brewer at a 7-barrel brewpub in Grand Rapids. I’ve pitched yeast that was too old, too cold, and too little, and I’ve logged the gravity readings to prove it. Every starter I make gets its OG measured before pitching and its cell count estimated from the growth curve, not guessed. I’ve killed two stir bars, one flask, and one cheap pump that died mid-sparge on batch 41. The gear below is what survived.

How I picked

My criteria were simple: does it grow enough cells for a 5-gallon batch of 1.060 beer, does it survive repeated boiling and cold crashing, and does it cost less than a ruined batch of ingredients? I tested three methods across 12 batches last year: a stir plate with a 2-liter flask, a shaken starter in a 1-gallon jug, and a no-stir starter that I swirled whenever I walked past the fermentation fridge. I also checked buyer reports on the stir plates that fail after six months and the flasks that crack in boiling water, because those are the two failure modes that waste a brew day.

Our pick: Stir plate with a 2L Erlenmeyer flask

Start with a used 2-liter Pyrex flask from a lab supply auction for $25, or pay $45 new. Pair it with a stir plate that has a brushless motor, not the $30 ones with a fan motor that dies at batch five. I run a 1.040 starter of DME for 24 hours and get a cell count that’s roughly double what a shaken starter produces. The proof is in the lag time: a stirred starter pitched into a 1.060 IPA shows activity in 8 hours. The same yeast, shaken, takes 14. That’s not opinion, that’s my brew log from batches 44 and 45.

The flask is the part that matters. A 2-liter flask with a foam stopper handles a 1.5-liter starter with room to spare. I’ve boiled starters in the same flask for three years and it’s still clear, no cracks, no cloudiness. Buyers on the homebrew forums report the same: the 2-liter size is the sweet spot, big enough for a lager starter, small enough to fit in most fridges for cold crashing.

The stir plate itself is where people cheap out and regret it. A $35 plate with a DC fan motor will stir fine for three months, then start stuttering at 500 RPM. The $80 plates with a brushless motor and a stronger magnet hold a 2-inch stir bar at 1,000 RPM for 48 hours straight. One buyer reported their cheap plate died mid-starter on batch two, leaving them with a half-stirred slurry and a stalled fermentation. That’s the exact failure I’m trying to prevent.

Flaws but not dealbreakers: The stir bar can jump the vortex if you run it too fast, flinging yeast onto the flask walls. It’s harmless, just annoying. The plate itself adds $80 to your setup. Over 40 batches, that’s $2 per batch for better attenuation and fewer stalled fermentations. Worth it.

This is for the brewer who logs gravity readings, wants consistent lag times, and plans to reuse yeast for more than one batch. If you’re on batch two and just want beer, skip this for now.

Budget pick: The shaken-not-stirred starter in a 1-gallon jug

You already own a 1-gallon glass jug from that cider kit or the pickle jar you cleaned out. Boil 1 liter of water with 100 grams of DME for 15 minutes, cool it to 70°F, pour in your yeast vial, cap it with foil, and shake it for 2 minutes every time you walk past. It grows about half the cells of a stirred starter, which is fine for a 5-gallon ale under 1.060.

I used this method for my first 20 batches. The beer was drinkable, the lag times were longer, and I had one batch of stout stall at 1.024 because I underpitched a 1.075 wort with a shaken starter that only grew 80 billion cells. That batch taught me the limit: this method works for session ales, not big beers.

The jug has one advantage over the flask: it’s cheap to replace. I cracked one by setting it on a cold counter right off the boil. A replacement costs $8 at the hardware store. The flask costs $45. For the budget brewer, that risk profile makes sense.

Flaws but not dealbreakers: The cell count is lower, so you’ll wait longer for fermentation to start. And you have to remember to shake it. I forgot twice, and both batches had sluggish starts. The beer survived, but the lag time stretched from 12 hours to 24.

This is for the brewer who wants to use liquid yeast for the first time without buying new gear. You’ll get better beer than pitching the vial directly, at zero cost.

Skip it: The no-stir starter

This is the method of swirling the jug once a day and hoping. It keeps yeast alive, but it doesn’t grow enough cells to make a difference. My batch 38 was a 1.065 amber ale pitched with a no-stir starter that had been sitting for 36 hours. It took 30 hours to show activity and finished at 1.018, which is 0.006 higher than the same recipe with a stirred starter. That’s the difference between a beer that’s crisp and one that tastes slightly sweet.

The only reason to use this method is if you have zero counter space and zero budget. It’s better than pitching a vial that’s been in the fridge for six months, but barely. The lag time is long enough that the yeast are more likely to produce off-flavors from the stress of starting in a high-gravity environment.

Flaws but not dealbreakers: There are no dealbreakers because I don’t recommend it. If you’re going to bother making a starter, shake it. The extra 2 minutes of effort doubles your cell count.

The competition

The 1-liter flask is too small for anything above a 1.040 starter. I bought one for $20, used it twice, and it sits in the “maybe sell” box. The 3-liter flask is too big for most fridges and costs $60. The stir plates under $50 all have the same fan motor that fails, based on the complaint pattern across three brands. The $150 lab-grade stir plates work, but you’re paying for a timer and a digital display you don’t need.

The one product I’ll never recommend is the “starter kit” that includes a plastic jug, a rubber stopper, and an airlock. The stopper never seals right, the jug cracks after three boils, and the airlock is useless because you don’t need one for a starter. You want oxygen in, not out. That kit is $25 of gear you’ll replace within a year.

The method that matters more than the gear

Boil 100 grams of DME per liter of water for 15 minutes. Cool it to 70°F. Pitch your yeast. Stir or shake. That’s it. The gear is just a way to get more oxygen into the wort, which is what the yeast need to build cell walls and reproduce. A starter made with tap water that has chlorine will stall your yeast before you even pitch it. Use filtered water or boil the tap water for 20 minutes to drive off the chlorine. I learned that after batch 22 tasted like band-aids.

The other variable is timing. A starter needs 18 to 24 hours to reach maximum cell density. If you pitch at 12 hours, you’re underpitching. If you wait 48 hours, the yeast have consumed all the sugar and start going dormant. My brew day schedule is: make the starter the night before, pitch it the next evening. That’s a 24-hour window that works every time.

Cost per batch for the stirred method: about $1.50 in DME and electricity. Cost per batch for the shaken method: $1.50 in DME and 4 minutes of shaking. The flask and stir plate pay for themselves after 10 batches if you’re buying fresh yeast vials at $10 each, because a starter lets you harvest and reuse yeast for 5 to 7 batches before it needs replacing.

That’s the real math. A $10 vial of yeast, used to build a starter, becomes $50 to $70 worth of yeast across multiple batches. The stir plate isn’t a luxury, it’s a way to stop paying $10 every time you brew.