How to count cells with a hemocytometer: the squares, the 10⁴ factor and trypan blue viability

Published 10 November 2026 · 19 min read

The tool for this guide: Open the multi counter

Short answer

Cells per mL equals the cells you counted, divided by the volume they were in, times the dilution. One large square of an improved Neubauer hemocytometer holds 0.1 µL, which is 10⁻⁴ mL, so the average per large square × 10⁴ × the dilution gives cells per mL. After a 1:1 mix with trypan blue, the dilution is 2.

The formula: cells counted ÷ volume counted

Every hemocytometer calculation is one division: cells per mL = cells counted ÷ volume counted (mL) × dilution factor. The World Health Organization's semen manual states it in those terms, the number "divided by the volume in which they were found ... multiplied by the dilution factor" (WHO laboratory manual for the examination and processing of human semen, 5th edition, 2010, section 2.8.4). The familiar 10⁴ is that division done in advance for one particular square.

The volume comes from the chamber. A hemocytometer is a thick slide with a ruled grid and a fixed gap between the grid and a special coverslip. On the standard improved Neubauer chamber the gap is 0.1 mm and the grid is nine squares of 1 mm by 1 mm, according to the manufacturer Marienfeld's counting-grid specifications. So one large square holds 1 mm × 1 mm × 0.1 mm = 0.1 mm³ = 0.1 µL = 10⁻⁴ mL.

A count in 10⁻⁴ mL is multiplied by 10⁴ to give a count per mL. That is why Thermo Fisher's Gibco protocol, Counting cells in a hemacytometer, counts the large central square and multiplies by 10⁴, and why STEMCELL Technologies' method writes it as the average cell count per square × dilution factor × 10⁴.

Why Thermo Fisher gives both 10⁴ and 10³

The two factors are the same rule applied to different volumes. Thermo Fisher's counting page has a second formula, C = N × D × 10³, where N is the viable cells counted "in 10 subgrids (1.0 mm³)" and D the dilution. Ten large squares at 0.1 µL each hold 1 µL, which is 10⁻³ mL, so a total over ten squares is multiplied by 10³. A count from one square is multiplied by 10⁴.

The rule extends to any set of squares, and it is the way to check any factor you are handed:

The factor for each set of squares on an improved Neubauer chamber, from 1 ÷ volume
What you counted	Volume counted	Multiply the total by	Or the average per square by
1 large square (1 mm²)	0.1 µL = 10⁻⁴ mL	10,000	10,000
4 large corner squares	0.4 µL	2,500	10,000
5 large squares (corners and centre)	0.5 µL	2,000	10,000
All 9 large squares of one chamber	0.9 µL	1,111	10,000
10 large squares, for example 5 in each of two chambers	1 µL = 10⁻³ mL	1,000 (Thermo's 10³)	10,000
1 group square (0.2 mm × 0.2 mm)	0.004 µL	250,000	250,000
5 group squares in the centre square	0.02 µL	50,000	250,000
1 smallest square (0.05 mm × 0.05 mm)	0.00025 µL	4,000,000	4,000,000

Multiply the result by the dilution factor in every row. The mistake the table guards against is mixing the two columns: the total over four squares multiplied by 10⁴ is four times too high, and the total over ten squares multiplied by 10⁴, instead of Thermo's 10³, is ten times too high.

The grid, square by square

The improved Neubauer grid is 3 mm by 3 mm, divided into nine large squares of 1 mm². Marienfeld's specification divides the central large square into 25 group squares of 0.2 mm by 0.2 mm, each split again into 16 smallest squares of 0.05 mm, 400 in all. The squares are separated by triple lines, with the central line as "the actual dimension line", and Marienfeld notes that leucocytes are counted in the four large corner squares.

One large square of an improved Neubauer chamber holds 0.1 µL, so each cell counted in it stands for 10,000 cells per mL of what was loaded; the smaller squares in the centre hold less, and each cell in them stands for more.

Two older or rarer rulings look similar and are not the same. Marienfeld's plain Neubauer (not improved) divides its central square into 16 group squares, not 25. Its Thoma chamber has a single 1 mm² ruled area with no large squares around it. Large squares on the improved Neubauer, Bürker and Thoma chambers all hold 0.1 µL; the difference is only in how they are subdivided.

Other chambers: Bürker, Fuchs-Rosenthal, Nageotte

Calculator pages list fixed "chamber factors" for Bürker and Thoma chambers without saying which squares they assume, so they cannot be checked. Work the factor out from the chamber's depth and the area you counted instead. These are the dimensions Marienfeld publishes:

Depth, squares and volumes of common counting chambers (Marienfeld)
Chamber	Depth	Squares	Volume of one square	One cell in it equals
Improved Neubauer	0.1 mm	9 of 1 mm²; centre into 25 of 0.04 mm² and 400 of 0.0025 mm²	0.1 µL; 0.004 µL; 0.00025 µL	10⁴; 2.5 × 10⁵; 4 × 10⁶ per mL
Bürker	0.1 mm	9 of 1 mm², each with 16 squares of 0.04 mm²	0.1 µL; 0.004 µL	10⁴; 2.5 × 10⁵ per mL
Thoma	0.1 mm	One 1 mm² area of 0.04 mm² group squares and 0.0025 mm² squares	Whole area 0.1 µL (calculated); 0.004 µL	10⁴ for the whole area; 2.5 × 10⁵ per mL
Fuchs-Rosenthal	0.2 mm	16 of 1 mm², each with 16 of 0.0625 mm²	0.2 µL; 0.0125 µL	5,000; 80,000 per mL
Nageotte	0.5 mm	40 strips of 0.25 mm × 10 mm (2.5 mm²)	1.25 µL (calculated)	800 per mL

The deep, large chambers are for sparse suspensions. The whole Fuchs-Rosenthal grid holds 3.2 µL, 32 times a Neubauer large square, so a few cells per square still add up to a countable total. The arithmetic is identical; only the volume changes.

How to count, step by step

The steps below follow Thermo Fisher's and STEMCELL's protocols for cultured cells, with the chamber-handling details from the WHO semen manual, which is the most explicit published procedure.

Clean and seat the coverslip. Clean the chamber and coverslip (70% ethanol in STEMCELL's protocol) and dry them. Use the thick coverslip made for the chamber: the WHO manual specifies thickness number 4, 0.44 mm, and treats iridescent Newton's rings between the glass surfaces as the sign it is seated properly.
Mix the cells with stain. For viability, mix equal volumes of the suspension and 0.4% trypan blue, such as 10 µL of each, in serum-free solution. Strober's protocol in Current Protocols in Immunology (2015) explains why serum is left out: serum proteins take up the dye.
Load about 10 µL. Touch the tip to the edge of the coverslip and let capillary action fill the chamber. Do not overfill it, which lifts the coverslip and deepens the chamber, or underfill it, which leaves air in the grid (WHO 2.8.2).
Count at once. Focus at 10× and count. Strober says to count within 3 to 5 minutes of mixing, since longer incubation kills cells and lowers viability; Thermo says examine immediately.
Count whole squares in a fixed order. Count clear cells and blue cells separately, square by square, zig-zagging along the rows so no square is counted twice. The WHO manual adds a rule worth keeping for any count: finish the row or square you are on, never stop in the middle of one.
Count both chambers. A hemocytometer has two grids. Load the second from a separate mix and use it as the replicate check below.

Simple Online Counter is a free, browser-based set of counters that keeps its counts in your browser's own storage. In the multi counter you can name one counter "clear" and one "blue", or one per square, tap as you go, undo a double tap and export the counts as a CSV. It does not recognise cells; deciding what is a cell, and whether it is blue, stays with you at the microscope.

Which cells on the lines to count

Count the cells that touch two adjacent sides of each square and skip those on the other two. Each line then belongs to exactly one of the two squares it separates, so no cell is counted twice and none is missed. Which pair you choose matters less than using it every time, and the published methods do not agree on the pair, as ChemoMetec's guide also notes:

The edge rule in three published methods
Source	Count cells on	Skip cells on
WHO semen manual, 5th edition (2010), section 2.7.3	Lower and left lines, an "L" shape	Upper and right lines
STEMCELL Technologies	Bottom and left borders	Top and right borders
ChemoMetec	Top and right borders	Bottom and left borders

On a chamber with triple lines, Marienfeld and the WHO manual both take the middle line as the boundary. The WHO manual counts a spermatozoon if most of its head lies between the two inner lines, and not if most of it lies between the two outer lines. A cell sitting squarely on the middle line is the one the L-rule decides.

Under the lower-and-left rule this square counts 7 cells: the 4 inside it, 2 on its left line and 1 on its lower line; the 3 cells on its top and right lines are counted with the neighbouring squares instead.

Worked example 1: trypan blue, four corner squares

A culture is mixed 1:1 with 0.4% trypan blue and 10 µL is loaded. The four large corner squares hold:

Counts from the four corner squares of one chamber
Square	Clear (live)	Blue (dead)
Top left	52	6
Top right	47	4
Bottom left	55	7
Bottom right	50	5
Total	204	22

Volume counted. Four large squares × 0.1 µL = 0.4 µL = 0.0004 mL.
Live cells per mL in the chamber. 204 ÷ 0.0004 = 510,000.
Undo the dilution. The 1:1 mix halved the concentration, so × 2 = 1,020,000 live cells per mL, reported as 1.0 × 10⁶. The short route gives the same: the average of 51 per square × 2 × 10⁴.
Total cells. (204 + 22) = 226 ÷ 0.0004 × 2 = 1,130,000 per mL, or 1.1 × 10⁶.
Viability. 204 ÷ 226 × 100 = 90.3%, using Thermo Fisher's formula [1 − (blue ÷ total)] × 100, which comes to the same thing.

Thermo Fisher's trypan blue exclusion protocol expects at least 95% viability for a healthy culture in log phase, so 90.3% is worth a note. To seed 200,000 live cells, take 200,000 ÷ 1,020,000 = 0.196 mL, about 196 µL, of the original suspension. The tally percentage calculator turns the clear and blue counts into percentages if you would rather not do it by hand.

Worked example 2: small, dense cells in the centre square

Dense suspensions of small cells overcrowd a large square, so they are counted in the smaller squares of the centre. Marienfeld's guide says erythrocytes are normally counted in at least five group squares. Suppose a blood sample is diluted 1:200 and 460 cells are counted in the four corner group squares and the middle group square of the central large square.

Volume counted. Five group squares × 0.004 µL = 0.02 µL.
Cells per µL in the chamber. 460 ÷ 0.02 = 23,000.
Undo the dilution. 23,000 × 200 = 4,600,000 per µL, which is 4.6 × 10¹² per litre, since a litre is a million microlitres.

The same counts in cells per mL would be 4.6 × 10⁹. Haematology reports per microlitre or per litre, cell culture per millilitre; the count is the same and only the volume unit changes. This is the manual method only; clinical blood counts are made by analysers under a laboratory's validated procedure, and their reference ranges belong to that laboratory. On a stained smear, platelets can be counted against red cells without a chamber, by the red cell ratio method.

How many cells to count, and the check that two counts agree

Count at least 200 cells for a result you will act on, and 400 across two chambers for a good one. Cells settle at random, so a count follows a Poisson distribution, in which the standard error is the square root of the count. The WHO semen manual (Box 2.7 and Table 2.2) turns that into the sampling error as a percentage of the count:

Sampling error by total cells counted (WHO semen manual, 2010, Table 2.2)
Cells counted	Sampling error	Approximate 95% range
25	20%	± 40%
50	14.1%	± 28%
100	10%	80 to 120
200	7.1%	172 to 228
400	5%	360 to 440
500	4.5%	± 9%

That is only the counting error. Pipetting, mixing and filling the chamber add their own, which is why the WHO manual counts two independent dilutions and compares them before accepting either. Its test (Box 2.10) works for any cell count: the difference between two counts of the same volume should be no more than 1.96 × √(their sum). A larger gap points to miscounting, a pipetting error or cells that were not well mixed.

Continuing worked example 1: the second chamber, loaded from a separate trypan blue mix, gives 230 clear cells in the same four squares.

The sum is 204 + 230 = 434, and the difference is 26.
1.96 × √434 = 1.96 × 20.8 = 40.8. The WHO table rounds the limit for sums of 427 to 448 to 41.
26 is under 41, so the counts agree. The mean is 217 per 0.4 µL: 217 ÷ 0.0004 × 2 = 1,085,000, reported as 1.1 × 10⁶ live cells per mL, with a counting error of about √434 ÷ 434 = 4.8%.

Had the second chamber given 260, the sum would be 464 and the difference 56, against a limit of 42. The WHO rule is then to discard both counts and make two fresh dilutions, not to count a third chamber and keep the two closest.

The dilution decides whether you can reach those totals. STEMCELL suggests a dilution that puts about 50 to 100 cells in each large square, which over four squares gives 200 to 400. At the 10⁴ factor that is 5 × 10⁵ to 1 × 10⁶ cells per mL in the chamber. A crowded square is hard to count without missing cells, and a sparse one needs many squares to reach 200.

Mistakes that skew a hemocytometer count

Most of the pages that rank for this question agree on the method. Where a count goes wrong, it is usually one of these:

Common hemocytometer errors and what the sources say
Mistake	Why it is wrong	Fix
Total of four squares × 10⁴	10⁴ is for one square's volume	Use the average, or the total ÷ 0.0004 mL
Thermo's 10³ on one square's count	10³ is for ten squares, 1 µL	Divide by the volume you actually counted
Forgetting the trypan blue dilution	A 1:1 mix halves the concentration	Multiply by 2 (Strober)
Counting cells on all four lines	Cells on shared lines are counted twice	Count two adjacent sides only
Leaving cells in trypan blue	Viability falls after 3 to 5 minutes (Strober)	Count at once
Staining in medium with serum	Serum proteins take up the dye (Strober)	Use PBS or serum-free medium
Overfilling the chamber	The coverslip lifts and the depth is no longer 0.1 mm	Reload a clean, dry chamber
Two counts from one dilution	It cannot detect mixing or pipetting errors (WHO)	Load each chamber from its own dilution
Reporting 1,085,000 cells per mL	Four digits the count cannot support	Two significant figures: 1.1 × 10⁶

What this guide does not cover

This guide covers manual counts of cells in suspension with a counting chamber, for teaching and research laboratories. Clinical counts on blood, cerebrospinal fluid or semen are made under a laboratory's validated procedure, often with a dilution fluid and chamber that procedure specifies, and are interpreted by the laboratory and the clinician against their own reference ranges. Nothing here is medical advice.

Two limits on the sources. The WHO manual quoted is the fifth edition of 2010; a sixth edition was published in 2021 and we could not open it, so check its wording if you work to it. ISO 20391-1, the international standard on cell counting methods, is sold rather than published and was not read for this page. For counting colonies on agar rather than cells in a chamber, see how to count bacterial colonies, and for classifying white cells on a blood film, the manual differential guide. A white count from the chamber multiplied by the differential's neutrophil fraction, bands included, gives the absolute neutrophil count. A total white count is half the picture; the normal differential ranges give the share expected for each cell type.

Frequently asked questions

What is the formula for counting cells with a hemocytometer?

Cells per mL = cells counted ÷ volume counted in mL × dilution factor. On an improved Neubauer chamber each large 1 mm square holds 0.1 µL, or 10⁻⁴ mL, so the everyday version is the average count per large square × 10⁴ × dilution. Four squares averaging 18.25 at a 100-fold dilution give 18.25 × 100 × 10⁴ = 1.8 × 10⁷ cells per mL.

Why do you multiply by 10,000 when counting cells with a hemocytometer?

Because one large square of the standard chamber is 1 mm by 1 mm with a 0.1 mm gap under the coverslip, a volume of 0.1 mm³. That is 0.1 µL, or one ten-thousandth of a millilitre. A count in one ten-thousandth of a millilitre is multiplied by 10,000 to give the count in a whole millilitre. The 10⁴ is a volume conversion, not a constant of nature.

Why does one formula use 10³ instead of 10⁴?

Because it counts ten large squares, not one. Thermo Fisher's formula C = N × D × 10³ takes N as the cells in 10 squares, which hold 1 mm³, or 1 µL, or 10⁻³ mL. Ten squares hold ten times the volume of one, so the factor is ten times smaller. Using 10³ on a single square's count gives an answer ten times too low.

Which squares do you count on a hemocytometer?

For cultured cells, usually the four large corner squares, or the four corners and the centre. The large squares are 1 mm across and suit cells that are not too dense. For small, dense cells such as red blood cells, Marienfeld's chamber guide says to count at least five of the 0.2 mm group squares inside the central square. Whatever you count, divide by the volume of exactly those squares.

Do you count cells touching the lines of a hemocytometer?

Count the cells on two adjacent sides of each square and skip those on the other two. The WHO semen manual (2010) counts cells on the lower and left lines; STEMCELL Technologies says bottom and left; ChemoMetec says top and right. Any pair works if you keep to it, because each line is then counted once, by one of the two squares it divides.

How do you calculate cell viability with trypan blue?

Viability is the unstained cells divided by all the cells counted, times 100. Live cells keep the dye out and stay clear; dead cells take it up and turn blue. With 204 clear and 22 blue cells, viability is 204 ÷ 226 × 100 = 90.3%. Thermo Fisher expects at least 95% for a healthy culture in log phase.

What is the dilution factor when you mix cells 1:1 with trypan blue?

Two. Equal volumes of cell suspension and 0.4% trypan blue, such as 10 µL of each, halve the cell concentration, so the count in the chamber is multiplied by 2 to get back to the original suspension. Strober's protocol in Current Protocols in Immunology does this explicitly, for both the viable count and the total count.

How many cells should you count on a hemocytometer?

At least 200 in total if the result matters, and more for precision. Counts follow a Poisson distribution, so the error is about 1 ÷ √N of the count: 10% at 100 cells, 7% at 200 and 5% at 400. The WHO semen manual asks for at least 200 per replicate and two replicates. STEMCELL suggests a dilution giving 50 to 100 cells per large square.

How long can cells sit in trypan blue before counting?

A few minutes at most. Strober's protocol says to count within 3 to 5 minutes of mixing, because longer incubation kills cells and lowers the viability count; Thermo Fisher says to examine the chamber immediately. The mix should also be in serum-free solution, since serum proteins take up the dye and confuse the reading.

How do you use a Fuchs-Rosenthal or Bürker chamber?

The same way, with that chamber's volumes. A Bürker large square is 1 mm² at 0.1 mm deep, 0.1 µL, exactly like a Neubauer large square, so the factor is again 10⁴. A Fuchs-Rosenthal chamber is 0.2 mm deep, so its 1 mm² squares hold 0.2 µL and one cell per square is 5,000 per mL. Always check the depth printed on the chamber.