Engineering Guide · Vendor-Neutral

Shake Flask vs Microtiter Plate for Screening: Which Format Should You Use?

Published September 15, 2026 · 12 min read

Figure 1: A single shake flask holds one condition at 20 to 100 mL working volume with well-understood scale-up physics. A microtiter plate holds 24 to 96 conditions per plate at 0.1 to 5 mL per well, with optional online DO, pH, biomass, and fluorescence optodes.

Quick Verdict

Use a microtiter plate for the primary screen (48 to 384 conditions per plate, online DO/pH/biomass via optical sensors) and a shake flask for the confirmation run on the top 10 to 50 hits (larger volume, RAMOS-based OTR, well-characterised scale-up to stirred tanks). MTPs and flasks are not substitutes but consecutive stages of a screening funnel.

Key differences at a glance

Shake flask: 20 to 500 mL working volume, 1 condition per vessel, kLa 50 to 150 h−¹. Sacrificial sampling is easy and scale-up to stirred tanks is well-characterised.
Microtiter plate: 0.1 to 5 mL working volume per well, 24 to 384 wells per plate, kLa 20 to 1140 h−¹ depending on well geometry. Online DO, pH, biomass, and fluorescence via optical sensor patches.
Best for the primary screen: microtiter plate. 10 to 20× more conditions per operator-day.
Best for scale-up confirmation: shake flask. Larger volume, better-characterised physics, RAMOS-based OTR that transfers cleanly to bench stirred tanks.
Cost per data point: MTP wins by 5 to 10× on consumables, but requires a BioLector or similar reader ($100k to $250k capital) to unlock its full value.

Side-by-side comparison

Factor	Shake flask	Microtiter plate
Working volume per condition	10 to 500 mL (typically 20% of vessel)	0.1 to 5 mL per well
Conditions per operator-day	20 to 40 with hand sampling	200 to 800 with a Duetz or BioLector platform
Oxygen transfer (kLa)	50 to 150 h−¹ (standard 250 mL, 50 mL fill, 200 rpm)	Up to 800 h−¹ (96-DWP) or 1140 h−¹ (48-well Flowerplate)
Online monitoring	Optional (RAMOS for OTR/CTR; PreSens sensor spots for DO/pH)	Standard on BioLector (biomass, DO, pH, fluorescence, all wells)
Sacrificial sampling	Easy — withdraw 1 to 5 mL as needed	Limited — must dedicate wells to time points or use non-invasive optodes
Reproducibility (well-to-well or flask-to-flask)	CV 5 to 15% with matched fill volume and shaker	CV 5 to 25% (edge effects, evaporation, lid condensation without sandwich cover)
Scale-up predictability	Well-characterised via Büchs and Maier correlations; matches bench stirred tanks within 20%	Predictive only when MTP kLa is measured and matched (Flowerplate: yes; standard 96-well U-bottom: no)
Typical consumable cost per condition	$3 to $15 (flask + baffle + closure + media)	$0.20 to $2 (plate + cover + media aliquot)
Capital equipment	Rotary shaker ($3k to $15k)	Plate shaker ($5k to $20k) plus BioLector or reader ($100k to $250k) for online data

Values reflect typical ranges for microbial and yeast cultivations in industrial process development. Mammalian cell cultures use different geometries (T-flasks, 6-well and 24-well plates) with lower shaking speeds and different kLa expectations.

Shake flask in detail

The shake flask is the century-old workhorse of microbial process development. A baffled or unbaffled Erlenmeyer flask (25 to 5000 mL) is filled to 10 to 20% of nominal volume and mounted on an orbital shaker (25 or 50 mm shake diameter) at 150 to 300 rpm. Oxygen enters through the closure (cotton plug, membrane cap, or vented Duetz sandwich cover), and the shaking generates the surface renewal that drives OTR.

How it works

Shake flask oxygen transfer is governed by Büchs' correlations: kLa scales with rotational speed to the 1.16 power, fill volume to the negative 0.83 power, and shake diameter to the 0.38 power. For a standard 250 mL Erlenmeyer flask at 50 mL fill and 200 rpm on a 25 mm orbit, kLa lands in the 50 to 150 h−¹ range — sufficient for microbial fermentations up to about 10 to 20 g/L cell dry weight before OTR becomes limiting. Adding a Kuhner RAMOS device measures OTR, CTR, and respiratory quotient online in eight flasks in parallel, giving quantitative oxygen uptake rate data that transfers cleanly to bench stirred-tank scale-up.

Flasks are typically closed with cotton or foam plugs for microbial work, or with vented membrane caps for CHO and other mammalian cell cultures where sterility and CO₂ exchange matter more than air throughput. The Enzyscreen sandwich cover is the gold-standard closure for reproducible parallel flask work — it suppresses evaporation while allowing controlled gas exchange, and pairs with the Enzyscreen Growth Profiler for automated OD readout.

When the shake flask wins

Shake flasks win in three situations. First, when the downstream analytical readout consumes more than a millilitre per time point — spent media analysis, LC-MS peptide mapping, or protein purification from culture supernatant. Second, when the process needs sacrificial sampling for a growth curve or productivity assay across 8 to 24 conditions rather than 96 or 384. Third, when the goal is scale-up characterisation for a stirred tank: the Büchs and Maier flask correlations, combined with kLa or OTR measurements, transfer cleanly to bench bioreactor scale-up to bioreactor workflows.

Shake flasks are also the practical follow-up format for the top 10 to 50 hits from a microtiter plate screen. Once the primary MTP screen has ranked hundreds of clones or media formulations, the confirmation runs move to shake flasks to gather higher-fidelity growth curves, product titre by HPLC, and biomass by dry cell weight rather than scattered-light proxies. Nearly every published industrial workflow uses this two-step funnel: MTP for breadth, shake flasks for depth.

Microtiter plate in detail

The microtiter plate condenses parallel cultivation onto a standard SBS-format substrate (127.76 mm by 85.48 mm, ANSI/SLAS-1). Formats range from 6 wells (deep, 10 mL per well) through 24 wells (2 to 4 mL) and 48 wells (1 to 3 mL) to 96 wells (0.1 to 0.5 mL) and 384 wells (30 to 100 microlitres). Well geometry — round-bottom, U-bottom, V-bottom, square-bottom, or the proprietary Flowerplate shape from Beckman Coulter (m2p-labs) Flowerplate — strongly determines kLa and mixing behaviour.

How it works

Microtiter plate cultivation happens on a high-frequency plate shaker (typically 1000 to 1400 rpm at a 3 mm orbit for microbial work, or 150 to 300 rpm at a 50 mm orbit for mammalian). Oxygen enters through the plate cover or a gas-permeable film, and reaches the culture through the top surface of each well. Because the surface-to-volume ratio is much higher than a shake flask, kLa can be much higher — up to 1140 h−¹ in the 48-well Flowerplate at 1400 rpm and 800 h−¹ in 96-deep-well square plates. Standard 96-well U-bottom plates at 1000 rpm reach only 100 to 250 h−¹, which is below shake flask performance and often limits high-cell-density cultivations.

Online monitoring transforms the MTP from a hit-picking assay into a microbioreactor system. The Beckman Coulter (m2p-labs) BioLector XT measures biomass by scattered light, pH and DO by non-invasive PreSens optical sensor spots, and product fluorescence in all 48 wells of a Flowerplate. The BioLector Pro adds microfluidic feeding and pH control via microvalves on a disposable microfluidic MTP — enabling fed-batch and controlled-pH cultivations in a 32-well microbioreactor. Automated MTP handling platforms — the Duetz System for reproducible parallel MTP shaking, Beckman Coulter Biomek or Tecan Freedom EVO for liquid handling — push throughput to 400 to 800 conditions per operator per week.

When the microtiter plate wins

The MTP wins whenever the number of conditions to test exceeds what one operator can hand-sample in shake flasks — typically anything above 40 conditions per week. Concrete examples: primary clone screening after transformation (hundreds of transformants), DOE media optimisation with 40 to 128 runs, IPTG induction sweeps across strain and inducer concentration, or high-throughput definitive screening designs for early process development. When online DO, pH, and biomass are available, MTPs also support quantitative physiological screening rather than just endpoint hit-picking.

The MTP also wins on cost per data point once you own a BioLector or comparable reader. Consumables per condition drop from $3 to $15 in a shake flask (glass or PETG flask, closure, media aliquot) to $0.20 to $2 per well in a plate (disposable plate, gas-permeable seal, media aliquot). Amortised across a strain engineering campaign that runs thousands of conditions per year, the MTP payback on a $150k BioLector Pro is typically 6 to 18 months versus a shake-flask-only workflow.

Pros and cons

Shake flask

Advantages

Well-characterised physics via Büchs correlations — kLa and OTR predict cleanly from geometry, fill, and shaker settings.
Sacrificial sampling is trivial — 1 to 5 mL withdrawn per time point for any offline assay.
Transfer to stirred-tank bioreactor is decades-established, with matched kLa or OTR the standard scale-up criterion.
Low capital cost — a $3k to $15k orbital shaker runs 8 to 24 flasks in parallel.

Disadvantages

Throughput is limited — 20 to 40 conditions per operator-day is a hard ceiling with manual sampling.
Online monitoring requires bolt-on hardware (RAMOS, PreSens sensor spots, Enzyscreen Growth Profiler) that adds $15k to $100k.
Consumable cost per condition is 5 to 10× higher than a well of a microtiter plate.
Not practical for the very high condition counts required by modern strain engineering and DOE workflows.

Microtiter plate

Advantages

24 to 384 conditions per plate; automatable to 400 to 800 conditions per operator per week with a Biomek or Tecan platform.
Online DO, pH, biomass, and fluorescence in all wells via BioLector or Reader-compatible sensor plates.
Very high kLa possible with square 96-deep-well or 48-well Flowerplate geometries (up to 1140 h−¹).
Consumables per condition are 5 to 10× cheaper than shake flasks, driving down cost of DOE and screening campaigns.

Disadvantages

Standard 96-well U-bottom plates deliver less oxygen than shake flasks — not suitable for high-cell-density work without deep-well or Flowerplate geometry.
Sacrificial sampling is impractical — must dedicate wells to time points or rely on non-invasive optodes only.
Edge effects, evaporation, and lid condensation can push CV above 15% without a sandwich cover and humidified chamber.
Filamentous fungi, high-foaming strains, and cultures that need physical scale-up all show artefacts that a shake flask would not.

Which should you choose?

Match the format to the stage of the screening funnel and the type of readout you need.

Primary strain or media screen (100 to 1000 conditions)

Rank hundreds of clones, media formulations, or DOE runs by biomass and product titre. Online DO/pH/biomass optional but strongly preferred for physiological insight.

Choose microtiter plate (Flowerplate or 96-DWP)

Confirmation of top hits (10 to 50 conditions)

Higher-fidelity growth curves, sacrificial sampling for spent media and HPLC, and Büchs-correlation-based OTR that transfers to stirred tanks.

Choose shake flask (250 mL to 500 mL)

Scale-up to bench stirred tank (top 3 to 5)

Match OTR from RAMOS-instrumented shake flasks to bench-scale kLa via constant-kLa or constant-vs scale-up. MTP data alone is rarely sufficient.

Choose shake flask + RAMOS characterisation

Budget-constrained lab (<$50k capex)

No BioLector or Biomek available. Shake flasks with a Growth Profiler or RAMOS deliver most of the screening value at a fraction of the capital cost.

Choose shake flask (Enzyscreen + Growth Profiler)

Real-world use cases

Typical setups where bioprocess teams have converged on one format or the other.

E. coli protein expression

Flowerplate on BioLector for the primary screen

Primary screen of 48 to 96 clones for IPTG induction response — biomass, DO, and eGFP fluorescence read continuously. Top 12 clones move to 250 mL shake flasks with RAMOS OTR before 5 L bench bioreactor.

Yeast media DOE

96-DWP with Duetz sandwich covers

Response surface DOE with 40 to 60 runs on media composition. Square 96-DWP at 1000 rpm hits kLa 400 h−¹, matching bench yeast fermentation. Endpoint OD and titre by HPLC.

CHO clone screening

24-well deep plate on ambr15 or shake flask

CHO clone development typically uses a lower-shear 24-well deep plate at 150 to 300 rpm on a 25 mm orbit — kLa 20 to 50 h−¹, matched to shake flask follow-up before ambr250 or 3 L bench.

Filamentous fungi

Shake flask remains the practical option

Filamentous cultures (Aspergillus, Trichoderma, Streptomyces) show wall growth and morphology artefacts in most MTPs. Baffled shake flasks with RAMOS remain the standard for the primary screen through to scale-up.

Not sure how your screen conditions scale to a stirred tank?

The Scale-Up Calculator applies P/V, kLa, tip speed, Reynolds and mixing-time criteria side by side, so you can convert a shake flask OTR or a BioLector Flowerplate kLa into an impeller speed at bench, pilot, or manufacturing scale.

Open the Scale-Up Calculator

Cost and lifecycle considerations

Total cost of screening is capex plus consumables plus time

A shake-flask-only workflow costs almost nothing to start ($5k to $15k for a shaker and glassware) but caps throughput near 30 conditions per operator per day. A BioLector-driven MTP workflow costs $100k to $250k upfront but drops per-condition cost by 5 to 10× and increases parallel throughput by 10 to 20×. For any strain-engineering or DOE campaign screening more than about 1,500 conditions per year, the BioLector pays back within 12 to 18 months.

The other cost that shifts is failed screening time. Standard 96-well U-bottom plates without a sandwich cover show CV of 15 to 25% due to edge effects and evaporation. A campaign of 1,000 conditions with that variance loses roughly 20 to 30% of its hits to noise, versus under 10% with a Flowerplate or Duetz-covered plate. In strain engineering, that translates to real programme delay: a lost hit is a strain that never makes it to scale-up. Investing in the reproducibility infrastructure (sandwich covers, humidified chamber, dedicated MTP shaker) is often cheaper than accepting the noise.

Shake flask consumables (glass or single-use PETG) run $3 to $15 per condition including closure and media aliquot. Microtiter plate consumables (Flowerplate, gas-permeable seal, media aliquot) run $0.20 to $2 per well. For a 400-run DOE, the shake flask consumables come to $1,200 to $6,000; the MTP consumables come to $80 to $800.

Cost component	Shake flask	Microtiter plate
Shaker / reader capex	$3k to $15k orbital shaker	$5k to $20k plate shaker; $100k to $250k with BioLector
Consumable per condition	$3 to $15	$0.20 to $2
Conditions per operator-day	20 to 40 (manual)	200 to 800 (automated MTP)
Time-to-first-result	Same day — standard glassware	2 to 4 weeks of setup and validation for a BioLector workflow

Vendor and platform landscape

The screening market is split between shake flask instrumentation (RAMOS, Growth Profiler, sensor spots) and full microbioreactor platforms (BioLector, ambr15, Micro-24). Each has a distinct sweet spot.

Shake flask platforms

Kuhner RAMOS: The reference standard for online OTR, CTR, and RQ measurement in shake flasks. 8 parallel flasks with in-line off-gas analysis — the go-to instrument for scale-up-relevant physiological screening.
Enzyscreen Growth Profiler: Automated OD readout for 96 microtiter plates, or the Enzyscreen sandwich cover for reproducible parallel shake flask work with suppressed evaporation and controlled gas transfer.
PreSens sensor spots: Non-invasive optical DO and pH measurement in shake flasks, single-use bags, and MTPs — the enabling technology for online monitoring in any transparent vessel.
Corning, Thermo Fisher Nunc: Single-use PETG shake flasks with vented caps — the workhorse consumable for microbial and mammalian shake-flask work.

Microtiter plate platforms

Beckman Coulter m2p-labs BioLector Pro / XT: 48-well Flowerplate with online biomass, DO, pH, and fluorescence. BioLector Pro adds microfluidic feeding and pH control for 32 fed-batch microbioreactors. The category-defining product.
Sartorius ambr15 / ambr250 modular: Stirred, sparged, sensor-instrumented microbioreactors in a 24- or 48-vessel cassette. Higher fidelity than a Flowerplate but not truly a microtiter plate — each vessel has its own impeller, sparger, DO, and pH probes.
Duetz System: Sandwich covers and clamps for 24-, 48-, and 96-well plates that suppress evaporation and provide reproducible mass transfer — the reference geometry for microbial MTP work.
Tecan Freedom EVO, Beckman Coulter Biomek: Automated liquid handling for parallel MTP inoculation, sampling, and endpoint assays — the throughput multiplier that turns a BioLector from an instrument into a screening factory.

Frequently asked questions

Is a microtiter plate the same as a shake flask?

No. A shake flask is a single-vessel cultivation format (typically 25 to 5000 mL Erlenmeyer with a working volume of 10 to 20 percent) mounted on a rotary shaker. A microtiter plate is a multi-well cultivation format (24, 48, 96, or 384 wells; 100 microlitres to 5 mL per well) mounted on a plate shaker. Shake flasks give one condition per vessel; MTPs give 24 to 384 conditions in parallel, at the cost of much smaller working volumes and different oxygen transfer characteristics.

Which has better oxygen transfer, shake flask or microtiter plate?

A standard shake flask (250 mL Erlenmeyer, 50 mL fill, 200 rpm, 25 mm shake diameter) achieves kLa around 50 to 150 h−¹. A standard 96-well round-bottom microtiter plate at 1000 rpm reaches kLa of roughly 100 to 250 h−¹. Purpose-built 96-deep-well square plates at 1400 rpm on a 3 mm shake diameter reach kLa of up to 800 h−¹ (OTR up to 200 mmol/L/h). The 48-well Flowerplate geometry reaches kLa of around 1140 h−¹. In short: standard 96-well U-bottom plates transfer less oxygen than shake flasks; deep-well and Flowerplate geometries transfer far more.

Can I scale up directly from a microtiter plate to a stirred tank bioreactor?

Only when the microtiter plate is characterised for kLa and matched to the target vessel on constant oxygen transfer capacity. The Flowerplate with online monitoring (BioLector) achieves kLa comparable to bench-scale stirred tanks, and published work shows less than 20 to 30 percent scale-up divergence in biomass and product titre for filamentous fungi and yeast when the OTR is matched. Standard 96-well U-bottom plates without online DO monitoring are not scale-predictive; use them for hit ranking, not for quantitative process transfer.

How many conditions can I run in parallel with each format?

A typical bench with two orbital shakers holds 8 to 24 shake flasks in parallel (with hand sampling, this is the practical daily throughput for one operator). One 96-well microtiter plate runs 96 conditions; one 48-well Flowerplate on a BioLector Pro runs 32 controlled microbioreactors plus 16 reservoirs. Automated MTP platforms (Duetz System, Beckman Coulter Biomek) can process 4 to 8 plates in a single run, giving 400 to 800 conditions per operator per week versus 20 to 40 with shake flasks.

Is a microtiter plate less reproducible than a shake flask?

It depends on the plate and workflow. Well-to-well evaporation, edge effects, and lid condensation on standard MTPs can push CV up to 15 to 25 percent between wells. Purpose-built screening formats (Enzyscreen sandwich covers, Duetz System, or Flowerplate with humidity control) push CV back to under 10 percent, comparable to or lower than shake flasks for filamentous cultures. A published study of Trichoderma reesei found MTP variability significantly lower than shake flask variability for extracellular metabolite production. The rule of thumb: purpose-built screening MTPs are as reproducible as shake flasks when humidity, evaporation, and mixing are all controlled.

When is a shake flask still the right choice?

Shake flasks remain the standard for follow-up experiments that need larger working volume (for sacrificial sampling, spent media analysis, or protein purification) and for well-characterised physical scale-up to stirred tanks. They are also the practical choice when the analytical readout consumes more than 1 mL per condition, when the process needs pH or DO probes larger than 5 mm, or when the strain has known plate-format artefacts (surface growth in filamentous fungi, high foaming). Most industrial workflows use MTPs for the first-pass screen (hundreds of conditions), then shake flasks for the top 10 to 50 hits, then bench bioreactors for the top 3 to 5.

What is a BioLector and does it replace shake flasks?

The BioLector (Beckman Coulter m2p-labs) is an orbital shaker with integrated online optical monitoring of biomass (scattered light), pH, dissolved oxygen, and fluorescence in a 48-well Flowerplate. The BioLector Pro adds microfluidic feeding and pH control. It does not fully replace shake flasks: BioLector is a screening platform (32 to 48 parallel conditions with online data), whereas shake flasks remain the workhorse for follow-up runs, sacrificial sampling, and process transfer to stirred tanks. Most modern strain engineering workflows use BioLector for the primary screen and shake flasks for confirmation.

How do RAMOS and BioLector compare for screening?

RAMOS (Kuhner, Adolf Kuhner AG) measures oxygen transfer rate (OTR), carbon dioxide transfer rate (CTR), and respiratory quotient in shake flasks (8 flasks in parallel). BioLector measures biomass, pH, DO, and fluorescence in a 48-well Flowerplate (48 conditions in parallel). RAMOS gives OTR-based physiological data on the same format used later for scale-up; BioLector gives many more conditions with less physical volume. Published work by Wilming and colleagues at RWTH Aachen recommends using both in parallel: BioLector for the parameter sweep, RAMOS for validating the top hits on the scale-up-relevant format.

Resources and references

Wilming A, et al. (2015). Parallel use of shake flask and microtiter plate online measuring devices (RAMOS and BioLector) reduces the number of experiments in laboratory-scale stirred tank bioreactors. Journal of Biological Engineering 9:9. — the reference for using RAMOS and BioLector in parallel for scale-up-relevant screening.
Meier K, Klöckner W, Bonhage B, Antonov E, Regestein L, Büchs J. (2016). Correlation for the maximum oxygen transfer capacity in shake flasks for a wide range of operating conditions and for different culture media. Biotechnology and Bioengineering 113(6):1191–1201. — comprehensive kLa correlations for shake flasks from Fernbach to microtiter plate scale.
Antonov E, Wirth S, Gerlach T, Schlembach I, Rosenbaum MA, Regestein L, Büchs J. (2014). Improvement and scale-down of a Trichoderma reesei shake flask protocol to microtiter plates enables high-throughput screening. Journal of Bioscience and Bioengineering (via ScienceDirect). — shows MTP variability lower than shake flask variability for filamentous fungi when purpose-built formats are used.
Klein T, Heinzle E, Schneider K. (2018). Assessment of the scalability of a microtiter plate system for screening of oleaginous microorganisms. Applied Microbiology and Biotechnology 102(11):5147–5159. — quantitative scale-up study from 96-DWP to 5 L stirred tank; less than 20–30% divergence in biomass and lipid yield.