← Back
Phase

CFOP Cross Tutorial

Review & Prerequisites

You already know how to make a bottom cross — LBL teaches you to build a daisy on top first, then flip each petal down with a half-turn. CFOP Cross solves the same step (four white edges on the bottom, white down, side colors matched to their centers) but without the daisy — you plan all four edges directly on the bottom, in as few moves as possible.

This step assumes you know LBL cross (the LBL cross tutorial) and have read the CFOP method overview, which explains why the cross goes on the bottom. Notation: R / L / F / B / U with ' and 2 suffixes.

Goal: An Efficient Cross

Scrambled — four white edges scattered across the cube
Bottom cross — four white edges solved, side colors aligned

Done = four white edges on the bottom, white down, side colors matched to their adjacent centers. A good cross typically takes just ~8 moves — not one edge at a time, but planned together after seeing all four.

Principle: Why Skip the Daisy

Why no daisy? LBL's daisy method first places four white petals on top, then flips each down — an entire extra step. CFOP Cross places the edges directly on the bottom, bypassing the daisy entirely. Same result, 4–5 fewer moves.

Why the bottom? With the cross on the bottom, the unsolved top layer stays visible throughout F2L (corner-edge pairs inserted simultaneously). The cross sits at the bottom, undisturbed by later moves — it's locked in place.

Why efficiency matters? Fewer moves = less time. More importantly, the faster you complete the cross, the sooner you start spotting your first F2L pair — a good cross leaves room for lookahead. Aim for a reasonable ~8 moves, not the perfect 5.

Why planning beats execution? The hard part of the cross is thinking, not turning — during the 15-second inspection you work out the order and the moves for all 4 edges entirely in your head. Execution is just carrying out the plan, no pausing to think. That's why speedcubers are already scanning for F2L while still turning the cross — the cross was solved in their head before the first move.

Edge Placement: Three Position Groups

No daisy — so how do you place each white edge? Look at which layer it's in. Three layers (top / middle / bottom), three approaches. Middle and bottom edges both reduce to the top: bottom → middle → top, one layer at a time.

Top layer, white up

White facing up (~2 moves): U-turn the side color to its matching center first — U only turns the top layer, never touching your placed bottom edges, so the alignment is "free." Then half-turn that face (R2 / F2…); the edge flips 180° down to the bottom, white-down, side color flush against the center. If white faces sideways, flip it up first then proceed (~3 moves).

Middle layer, white sideways

Inserting directly from the middle into the bottom would disturb your cross. First turn one face (R / L / F / B) to eject the edge to the top (~3 moves) — this turn doesn't touch the bottom cross, so placed edges stay safe. Then place it using the top-layer method — reducing to the top.

Bottom layer, white down but misplaced

White facing down but in the wrong position — it's occupying the correct edge's slot. First extract it to the top or middle layer, reopening the slot (~3 moves). Then place using the corresponding layer's method — reducing to top/middle.

Planning: Four Steps

What do you do during those 15 seconds of inspection? Follow four steps: Find → Read → Order → Estimate. Order has two strategies: easy-to-hard (place the simplest edge first) and pairing (two edges solved together, saving moves). Pairing principle: every face turn moves 4 edges on that face — if two target edges share a face, solving one "freely" repositions the other, possibly closer to its target, saving the moves you'd spend moving it separately. Criterion: two edges on the same face or adjacent faces. The strategies don't conflict — pair first, then easy-to-hard for the rest. Don't chase the optimal solution; learn to see systematically.

The cross is a relative structure — not 4 independent edges placed separately, but 4 fixed slots (DB↔DF opposite, DL↔DR opposite). Method: identify slots → compare positions → order. Identify: each white edge's non-white sticker matches a same-color center — that's its home. 4 edges = 4 slots forming a cross on the D face. Compare: how far is each edge from its target slot (top = far, bottom = close)? Order: use this spatial relationship to choose — don't just pick randomly.

Example 1: Easy-to-Hard

Find: 4 edges across 3 layers — DB on top white-up (target D-B), DL in the middle white-sideways (target D-L), DF on top white-sideways (target D-F), DR on the bottom white-down (target D-R).

Read: DR is closest to target (already on the bottom) but facing wrong — must extract; DB is farthest (on top) but white-up is easiest to place; DL is in the middle, eject then solve from the top; DF is on top white-sideways, flip up first then align — like DL, ~3 moves.

Order: easy-to-hard — DB (white-up, simplest) → DL (middle, eject) → DF (white-sideways, flip first) → DR (bottom, extract — hardest).

Estimate: ~9 moves (DB 2 + DL 1 + DF 3 + DR 3) — reasonable.

  1. 4 edges, 3 layersUB2（DB align+flip）
  2. DB placedL（DL insert）
  3. DB · DL placedU'R'F（DF flip+place）
  4. 3 edges placedB2RB2（DR extract+place）
  5. Cross done
0 / 9

Example 2: Middle-layer Pair

Find: DL, DR on top (targets D-L, D-R), DB, DF in the middle white-sideways (targets D-B, D-F) — the two middle edges share a similar position, pairing candidates.

Read: DL is simplest (top white-up); DB and DF are both in the middle, both misaligned — solving one might set up the other (pairing opportunity).

Order: pairing — top two first (DL → DR), then the middle pair DB + DF.

Estimate: ~9 moves (DL 2 + DR 2 + DB 4 + DF 1) — note DF takes only 1 move.

Notice: after solving DB, DF takes only 1 move — DB's ejection carried DF into position. That's pairing saving moves: two middle edges solved together cost 2–3 fewer moves than solving them separately.

  1. 2 top + 2 middleU2L2（DL align+flip）
  2. DL placedFR'（DR flip+place）
  3. Top layer doneLUL'B（DB eject）
  4. DB placed (pair)F2（DF insert）
  5. Cross done
0 / 9

Pitfalls

The traps here are mostly LBL habits and under-planning — not technical difficulty, but mindset.

  • Sticking to the daisy: building a daisy then flipping each petal down adds a whole extra step. Plan edges directly on the bottom — break the LBL habit.
  • One edge at a time, no overview: starting without seeing all four edges' relationships leads to extra U-turns and missed pairings. Look at the full board before turning.
  • Chasing optimal: trying to find a 5–6 move optimal cross wastes your 15-second inspection. A reasonable ~8 move cross is enough — plan first, optimize later.
  • Missing pairing opportunities: two edges that can be solved together go unnoticed. Ask yourself during inspection: can these two edges be placed without interfering? (Same layer, same group, non-interfering → possible pairing.)
  • Spending all inspection time just finding edges: 15 seconds is barely enough to locate all 4 edges — no time left to read relationships or choose an order, so you end up solving blindly one at a time. Scan all 4 positions quickly (Find ≤ 5 seconds), then spend the remaining 10 seconds on Read / Order / Estimate.

Practice, Then F2L

This page teaches planning — how to read the four edges' relationships, choose an order, and spot pairings in 15 seconds. The actual per-edge technique is on the practice page — practice each position group (top / middle / bottom) until you can place any edge without looking at the tutorial. Next step: F2L — the heart of CFOP efficiency, inserting corner-edge pairs simultaneously.

Practice cross placement