TLDR: If you are asking should you keep a two-land hand in MTG, keep it only when those lands cast your early spells and the hand remains functional if land three arrives late. Mulligan when the colors fail, your first relevant play costs three or more, or one missed land drop leaves you unable to participate.
Two lands can be perfect, acceptable, or disastrous. The useful question is not simply how many lands you have. It is what those particular lands let you do over the first few turns. Evaluate the hand as a sequence of plays, then adjust for your deck, matchup, position on the play or draw, and the cost of taking another mulligan.
Judge the sequence, not the land count
Start by describing your likely first three turns without assuming a favorable draw. Which spells can you cast? Do they advance your plan, answer likely pressure, or help find another mana source? If your honest answer is “play two lands and hope,” the hand is much weaker than its land count suggests.
A functional two-land hand usually has three qualities: the lands produce the colors its early spells require, there is at least one meaningful play before three mana, and the hand can absorb a delayed third land without becoming irrelevant. Wizards’ published mulligan guidance similarly treats land count as a baseline rather than the entire decision, pointing players toward color access, early plays, deck speed, matchup, and the cost of losing a card.
This is an example of a broader principle in improving your in-game decisions: make the plan and its assumptions visible. “This hand is probably fine” is hard to evaluate. “I can cast my turn-one selection spell, interact on turn two, and still function if I miss land three once” gives you a line that can be tested.
First, check colors and land timing
Two lands are not necessarily two usable mana sources. A hand with an early spell requiring one color and two lands that produce only another color is functionally missing mana. The same problem appears when both lands enter tapped and your hand depends on acting immediately.
Read the mana costs in order rather than in isolation. A hand might technically cast every spell eventually while still having no coherent curve. If your turn-one spell needs one color and your turn-two spell needs two sources of another, ask whether the exact lands in hand can support that sequence. Do not give a two-lander credit for colors you merely hope to draw.
Also account for restrictions attached to the lands. A colorless source may be excellent in the deck but unable to cast a colored early play. A tapped land may fix both colors but force you to choose which turn to delay. That delay can be harmless in a slower game and decisive when you must affect the battlefield early.
Use the third-land test
The cleanest test is simple: what happens if you do not draw land three by the turn you want to play it?
A resilient hand can still spend mana productively. It might deploy a cheap threat, hold up interaction, use card selection, or cast acceleration that is already supported by the two lands. A fragile hand contains several three- and four-mana spells with no useful action before them. Both hands have two lands, but only one is playing Magic while waiting.
Pay attention to whether the hand needs land three on time or merely benefits from it. If the entire plan collapses without the third land, you are making a high-risk keep. If land three improves an already functional sequence, the risk is much easier to accept.
Selection helps, but it is not another land
A cheap card-selection spell can justify a two-land keep because it gives you another look while using otherwise available mana. It does not guarantee the next land, however. The hand should have a reasonable fallback if the selection spell misses. It also needs the correct color to cast that spell in the first place.
Acceleration only counts when you can cast it
Treat a mana creature, artifact, or other ramp spell as part of the mana plan only if the opening lands cast it on schedule. A two-land Commander hand containing an affordable accelerator may have a credible route to four mana. A hand with two lands and a three-mana ramp spell still needs the third land before its “solution” works. That is dependency, not insurance.
Adjust for play, draw, matchup, and role
In a two-player game, the player taking the first turn skips that turn’s draw step. A borderline two-land hand is therefore less forgiving on the play: you see one fewer card before your early land drops than you would on the draw. That does not make every two-lander a mulligan on the play, but it should move close decisions toward caution.
The matchup matters because not every early action has equal value. Cheap removal may make a two-land hand excellent against an aggressive opponent, yet accomplish little against a strategy that develops through noncreature permanents. A slow value hand may be acceptable when both players have time, but unacceptable when failing to affect the first few turns will put you too far behind.
Your role matters too. If you need to apply pressure, a hand that cannot deploy a threat until turn three may fail even if its mana eventually works. If you need to survive, a hand containing castable interaction can be functional without a fast clock. The keep should support what your deck needs to accomplish in this game, not merely contain recognizable cards.
Compare the hand with a realistic mulligan
Do not compare an imperfect seven with an imaginary perfect six. Compare it with the range of six-card hands your deck is likely to produce. A shaky but functional seven can be better than redrawing in a deck with limited mana fixing or many expensive cards. Conversely, a seven that cannot execute any early plan is not protected by card count; those extra cards may remain stranded in your hand.
Under the mulligan procedure, you redraw to your starting hand size and then put a number of cards equal to your counted mulligans on the bottom of your library. This lets you evaluate a fresh group of cards before choosing what to bottom, but repeated mulligans still reduce the resources you keep.
After one mulligan, the threshold changes. A mediocre seven may be worth replacing because the first redraw costs less than keeping a nonfunctional opener. A functional six deserves more respect because going lower costs another card. Preserve a hand that can execute a credible plan; do not keep an unusable one merely because another mulligan feels bad.
Representative two-land opening hands
| Hand pattern | Likely decision | Reason |
|---|---|---|
| Two correctly colored lands, a one-mana play, a two-mana play, and useful follow-up | Usually keep | The hand spends early mana and remains active while looking for land three. |
| Two correct lands, cheap interaction, and several expensive spells | Conditional keep | Good if the interaction buys meaningful time; risky if the matchup makes it ineffective. |
| Two lands that cannot produce an early spell’s required color | Usually mulligan | The raw land count hides a color failure. |
| Two lands with no castable spell below three mana | Usually mulligan | Missing one land drop leaves the hand doing nothing. |
| Two lands plus acceleration that those lands can cast promptly | Often keep | The acceleration creates a real sequence, assuming it is reasonably likely to remain useful. |
| Two lands plus acceleration that itself costs three mana | Usually mulligan | The ramp does not solve the immediate third-land problem. |
These are starting points, not automatic rulings. A matchup-specific answer, unusually low curve, land-heavy deck, or especially powerful payoff can change the decision. The purpose of the patterns is to expose what the hand is asking you to risk.
How the answer changes by format
Two-player Constructed
Your deck’s known curve and role should drive the decision. A low-curve deck may operate comfortably on two lands, while a hand from a mana-hungry strategy may need its third and fourth land drops on time. Sideboarded games also provide matchup information: judge whether the early spells in hand interact with what the opponent is actually doing.
Limited
Limited hands often ask whether the available colors cast the early portion of your deck and whether you can survive until your more expensive cards come online. A two-land hand with both colors and relevant early plays can be functional. One with only a single color, no early action, and several expensive spells carries multiple failure points at once.
Multiplayer Commander
Commander can provide more time, but “the game will be long” is not enough to make a hand functional. Check whether you can develop mana or interact before your commander becomes castable. Also consider whether your early play is robust at a multiplayer table: spending a card to answer one minor threat may not compensate for falling behind the other players.
The multiplayer mulligan rule is important here. In a multiplayer game, a player’s first mulligan does not count toward the number of cards that player must put on the bottom afterward. In ordinary multiplayer Commander, that makes replacing a poor opening seven less costly. Use that opportunity to seek a functional hand rather than defending a two-lander whose only plan is drawing perfectly.
The table-side two-land checklist
Before keeping, answer these questions in order:
- Do these lands produce the colors my early spells actually require?
- Do tapped or restricted lands delay the sequence I am counting on?
- What can I cast before drawing a third land?
- If land three is one turn late, can I still make a meaningful play?
- Does my selection or acceleration work with the mana already in hand?
- Am I on the play or draw, and how many draw steps do I get before the critical turn?
- Does this hand perform my required role in the matchup or at this Commander table?
- Is this hand better than the realistic range of hands I would keep after a mulligan?
Keep when the colors work, the first turns contain meaningful actions, and a delayed third land is inconvenient rather than catastrophic. Mulligan when the colors fail, the hand cannot act before three mana, or every useful card depends on drawing one specific kind of mana source immediately.
Make the keep based on a real line
So, should you keep a two-land hand in MTG? Keep it when you can state a credible early sequence using only the resources already available, with reasonable draws improving that sequence rather than rescuing it. Send it back when “draw land three” is the entire plan.
On your next opening hand, do not stop at counting lands. Name your first meaningful play, check its colors, and ask what you do after a missed land drop. That ten-second test catches most of the two-land hands that look playable but are not.