What is CFOP? CFOP is a way to solve the 3x3 Rubik's Cube in 4 steps: Cross, F2L (the first two layers), OLL (orienting the last layer) and PLL (permuting the last layer). It is also called the Fridrich method, after Jessica Fridrich, who is credited with making it popular after she put her solving method online in 1997. It is the main method used in speedcubing competitions (Wikipedia).

Find your case fast

Look at your cube from above and tap the picture that matches it to see how to solve it. Grey cells are stickers you can ignore.

  • Case 1

    RUR′

    When the corner and the edge are apart in the top layer, R U R' both pairs them up and puts the pair into the front-right slot:

    1. You start with (input)
    2. After the formula
    See it in the post →
  • Case 2

    URU′R′

    When the pair is already joined and sitting right above its slot, U R U' R' moves the pair out of the way and then puts it in:

    1. You start with (input)
    2. After the formula
    See it in the post →
  • Case 3

    FRUR′U′F′

    F R U R' U' F' from the "Yellow cross" step turns a flat yellow line into the yellow cross:

    1. You start with (input)
    2. After the formula
    See it in the post →
  • Case 4

    RUR′URU2R′

    R U R' U R U2 R', called Sune, twists the remaining three corners so the whole top face turns yellow:

    1. You start with (input)
    2. After the formula
    See it in the post →
  • Case 5

    R2URUR′U′R′U′R′UR′

    For example, the U-perm R2 U R U R' U' R' U' R' U R' cycles three top-layer edges while the four corners stay put:

    1. You start with (input)
    2. After the formula
    See it in the post →

If you can already solve the cube with the layer-by-layer beginner method and want to get faster, CFOP is almost always the natural next step. This post explains each step and suggests a path that won't leave you overwhelmed.

How is CFOP different from the layer-by-layer method?

The beginner layer-by-layer method splits the cube into 7 small steps, each dealing with just a few pieces. CFOP merges many of those steps:

  • Layer 1 + layer 2: instead of putting in the 4 first-layer corners and only then the 4 middle-layer edges, F2L joins one corner with one edge into a pair and puts them in together.
  • Layer 3: instead of 4 steps (Yellow cross, Yellow face, Put the yellow corners in place, Put the yellow edges in place), full CFOP needs only 2 algorithms: one OLL to make the whole top face one colour, and one PLL to put every top-layer piece in place.

The result is fewer moves and fewer pauses to look at the cube, so you solve noticeably faster.

The 4 steps of CFOP: algorithm counts and target times

The table below puts the whole CFOP method in one place. Each row links to its own post about that step.

StepWhat you doAlgorithmsTarget time (20-second solve)Full guide
CrossJoin the 4 white edges into a cross on the bottom face0, done by intuition2 to 4 secondsHow to do the CFOP cross
F2LPair up 4 corners and edges and put them into the 4 slots41 cases, start with intuitive F2L9 to 11 secondsThe 41 F2L cases
OLLMake the whole top face yellow57 (2-look version: 10)2 to 4 secondsThe 57 OLL algorithms
PLLPut the top-layer pieces in place21 (2-look version: 6)3 to 5 secondsThe 21 PLL algorithms

The algorithm counts come from Wikipedia: adding the 41 F2L cases, full CFOP has 119 algorithms (Wikipedia).

The time column comes from two examples by CubeSkills, Feliks Zemdegs's cubing website, for someone who averages 20 seconds. A solver with weaker F2L who knows full OLL and PLL might take 4, 11, 2 and 3 seconds for the four steps. A solver with smooth F2L who still needs several looks for the last layer might take 2, 9, 4 and 5 seconds (CubeSkills).

Two things to take from the table:

  • F2L is about half of the solve. It is the step most worth practising.
  • The same total time can be split differently. CubeSkills points out that your split depends on what you already know. Time each step over many solves to find out which one is slowing you down.

If you're not near 20 seconds yet, don't worry about these numbers. Here is an easier first goal: J Perm writes that with plenty of practice, the cross can take under 5 seconds (J Perm).

TRY IT RIGHT HERE

Four moves.
One at a time.

Tap each symbol and follow one piece. This is an exercise in watching the movement.

The 4 steps of CFOP

1. Cross

This is like the "White cross" step of the layer-by-layer method: join the 4 edges with a white sticker around the white centre, with the other sticker of each edge matching the centre of its side face. The difference is that the cross goes on the bottom face, and CFOP solvers try to plan the whole cross during the inspection time before the solve starts. According to Wikipedia, the cross can always be solved in 8 moves or fewer (Wikipedia).

  1. You start with (input)A scrambled cube
  2. Result you will get after finishing this stepWhite cross done

Many CFOP solvers build the cross on the bottom face from the start, instead of making it on top and flipping the cube over, to save time. How to plan it and where each edge can be is covered in how to do the CFOP cross; once you are comfortable, the x-cross post shows how to make the cross and the first F2L pair at the same time.

2. F2L: the first two layers

F2L (First Two Layers) is the heart of CFOP. Each time, you find a first-layer corner and the middle-layer edge that goes with it, join them into a pair in the top layer, then put the pair into its slot. Do this 4 times for the 4 slots and the first two layers are done.

!Diagram of the F2L step in the CFOP method Image: [Conrad Rider / Wikimedia Commons](https://commons.wikimedia.org/wiki/File:CfopF2L.svg), CC BY-SA 3.0

F2L has 41 cases (Wikipedia), but you don't need to memorise them. The best way to start is intuitive F2L: understanding why a pair gets joined and put in, rather than remembering algorithms. Most cases are built from a few short sequences that follow the same shape: open the slot, turn the top, close the slot.

When the corner and the edge are apart in the top layer, R U R' both pairs them up and puts the pair into the front-right slot:

See it in 3D, before and after
  1. You start with (input)
  2. After the formula

When the pair is already joined and sitting right above its slot, U R U' R' moves the pair out of the way and then puts it in:

See it in 3D, before and after
  1. You start with (input)
  2. After the formula

In both pictures the white face is on the bottom and the other three slots are not disturbed.

Start with intuitive F2L, look up each case in the 41 F2L cases, then speed up with advanced F2L tips.

3. OLL: orienting the last layer

OLL (Orientation of the Last Layer) makes the whole top face one colour. The full set has 57 algorithms. The shorter 2-look OLL needs only 10 algorithms: 3 to orient the edges (making the yellow cross) and 7 to orient the corners (Wikipedia).

Good news: you already know some of them. Two algorithms from the layer-by-layer method are both part of 2-look OLL.

F R U R' U' F' from the "Yellow cross" step turns a flat yellow line into the yellow cross:

See it in 3D, before and after
  1. You start with (input)
  2. After the formula

R U R' U R U2 R', called Sune, twists the remaining three corners so the whole top face turns yellow:

See it in 3D, before and after
  1. You start with (input)
  2. After the formula

See all 10 algorithms in 2-look OLL, and all 57 cases in the full OLL algorithms.

4. PLL: permuting the last layer

PLL (Permutation of the Last Layer) puts the top-layer pieces in place to finish the cube. The full set has 21 algorithms; 2-look PLL needs only 6 algorithms (Wikipedia).

Once again, you already have a head start: the A-perm and U-perm you use in step 6 "Put the yellow corners in place" and step 7 "Put the yellow edges in place" of the layer-by-layer method are both PLL algorithms.

For example, the U-perm R2 U R U R' U' R' U' R' U R' cycles three top-layer edges while the four corners stay put:

See it in 3D, before and after
  1. You start with (input)
  2. After the formula

Start with 2-look PLL, then work through the 21 PLL algorithms and how to recognise PLL from just two faces.

A suggested path into CFOP

Getting into CFOP doesn't mean memorising 100 algorithms at once. A sensible path:

  1. Solve reliably with the layer-by-layer method. You should be able to solve without looking at a guide before you switch to CFOP.
  2. Switch to 2-look. Replace the 4 last-layer steps with 2-look OLL and 2-look PLL, 16 algorithms in total, and you already know many of them.
  3. Get the hang of intuitive F2L. Solve slowly, look for the corner-edge pair and try to put it in without breaking the slots you've already finished.
  4. Practise the cross on the bottom. Get used to looking at the cube and planning the cross before the timer starts.
  5. Add full PLL, then full OLL, bit by bit. A few algorithms a day, starting with the cases that come up most.
  6. Practise look-ahead. While you work on one F2L pair, try to spot the next one so you don't have to stop.

Is CFOP different from the Fridrich method?

No. CFOP and the Fridrich method are the same way of solving, just with different names. Same four steps, Cross, F2L, OLL, PLL, and the same algorithms.

The difference is who the name gives credit to:

  • "Fridrich method" is named after Jessica Fridrich. She put her solving method online in the 1990s, and that is how a lot of people first found out about it. Wikipedia gives the year as 1997; SpeedSolving Wiki says 1995.
  • "CFOP" is made from the first letters of the four steps. It isn't tied to any one person.

According to SpeedSolving Wiki, Fridrich is often wrongly named as the only inventor of the method. In fact many people helped: the idea of F2L pairs came from René Schoof and others, while OLL and PLL were developed by Hans Dockhorn, Kurt Dockhorn and Anneke Treep in the Netherlands, separately from Fridrich, in the early 1980s (SpeedSolving Wiki). Wikipedia also says the method grew out of work by many cubers in the early 1980s, and that Fridrich's main contribution was the set of OLL and PLL algorithms (Wikipedia).

That is why the community slowly moved to the name CFOP. SpeedSolving Wiki notes that the debate over the name became widely known around 2008, and CFOP has been the more common name since then. Even so, many people still see "Fridrich method" as a correct name.

For you it's simple: whether a guide says "Fridrich" or "CFOP", you're reading about the same method.

Do record holders use CFOP?

Yes. Wikipedia lists many top cubers who use CFOP, including Max Park, Feliks Zemdegs and Yiheng Wang (Wikipedia). Yiheng Wang currently holds the 3x3 average world record at 3.51 seconds, as of 14 October 2026 according to WCA data (WCA API).

Videos

The J Perm channel has a series of beginner CFOP videos that are very easy to follow. Start with intuitive F2L:

Learn F2L in 6 minutes (Full Intuitive F2L Tutorial)Video · J Perm · YouTube

Then 2-look OLL and 2-look PLL:

Rubik's Cube: Easy 2-Look OLL Tutorial (Beginner CFOP)Video · J Perm · YouTube Rubik's Cube: Easy 2-Look PLL Tutorial (Beginner CFOP)Video · J Perm · YouTube

Practise CFOP with EasyRubic

In the EasyRubic app, the step-by-step guide has two levels: Basic, with the 7-stage layer-by-layer method, and Advanced, with CFOP, including the 2-look, F2L, OLL and PLL algorithm groups. You can play every algorithm on a 3D cube to see each move, and there is a search box to look things up quickly.

Get EasyRubic to start your CFOP journey.

Sources

Frequently asked questions

Is CFOP hard for beginners?

Jumping straight into full CFOP is quite a lot. But if you go from the layer-by-layer method, through 2-look OLL/PLL (16 algorithms) and then intuitive F2L, every step is manageable.

How many algorithms are in full CFOP?

57 OLL algorithms and 21 PLL algorithms, 78 in total. F2L has 41 cases but is usually done by intuition; counting F2L as well, it's 119.

How long should each CFOP step take?

It depends on your level. For someone who averages 20 seconds, CubeSkills gives two examples: 4, 11, 2 and 3 seconds, or 2, 9, 4 and 5 seconds for Cross, F2L, OLL and PLL. In both, F2L takes about half the time.

Are CFOP and Fridrich the same thing?

Yes. They are two names for the same method. The name Fridrich comes from Jessica Fridrich, who made it popular; the name CFOP is made from the four steps and is now used more often.

Should I start with F2L or OLL/PLL?

Switch to 2-look OLL/PLL first, because you already know many of those algorithms from the layer-by-layer method, then put your effort into intuitive F2L, the part of CFOP that needs the most practice.

This information is based on competition results owned and maintained by the World Cube Association, published at worldcubeassociation.org/results as of 28.09.2026.

Cover photo: Lembasts / Wikimedia Commons, CC BY-SA 3.0

EasyRubic Editorial

Friendly notes for anyone who wants to solve their first Rubik’s Cube.

Ready for your first turn?

Go from reading notation to each step of a solved cube.

Start with step 0

Learn the 3x3 step by step · 3x3 online timer