Gender in Palworld breeding, and the 44 Pals with skewed ratios

You always need one male and one female in the Breeding Farm — but which parent is which changes the child for exactly one pairing in the entire game. Everything else about gender in Palworld breeding is a catching problem, not a breeding problem, and 44 species make that catching problem harder than a coin flip.

Verified on patch 1.0.3. Every ratio on this page is read from MaleProbability in DT_PalMonsterParameter, and all 289 values in our dataset were re-checked against that table for this page with no discrepancies.

The short answer

Gender does two jobs in Palworld breeding, and only two.

1. It gates the pairing. A Breeding Farm needs one male and one female. Two of the same gender produce nothing at all — see two females below for what the game does about that.
2. It picks the species in one single case. The game's override table, DT_PalCombiUnique, carries 258 rows, and 256 of them set both parent genders to None — the table's way of saying "either way round". The two remaining rows are the same pair of species read in both directions.

That is the whole of it. Gender does not touch passives, Potentials, mutation, egg size, hatch speed or breeding speed, because no field in the breeding data is scoped to it. We go through that in the male-parent question, which is the most expensive wrong belief in this corner of the game.

The one pairing where gender changes the child

Katress and Wixen. Both directions are shipped as separate rows, and they give different children:

Read the table by the female column and the pattern is immediate: in both rows the child is a variant form of the female parent's species. Male Katress with a female Wixen gives you the Dark Wixen; flip the genders and you get the Fire Katress instead. That is a description of two rows rather than a law of the game — there are no other gendered rows for it to generalise over — but it is the fastest way to remember which way round to load the farm.

Both parents are ordinary 50/50 species, so if you catch a random pair of each you have no control over which child you get. To choose, you have to supply the gender deliberately: the Katress you put in the farm decides it. There are 2 species in the game that exist on the other side of this rule — Wixen Noct and Katress Ignis — and they are the only two whose cross-species route is gender-dependent.

245 Pals are exactly 50/50

Out of 289 species, 245 carry MaleProbability = 50. That is worth stating as a count rather than a percentage, because the rounded figure is what makes people assume the exceptions are a rumour. For those 245, "you are just unlucky" is the correct answer to a run of one gender. For the other 44 it is not, and telling someone their Mozzarina luck is variance is simply wrong.

Correction: no Pal in Palworld is 5% male

The 5% figure circulates as the bottom of the range, usually attached to Bellanoir. It is not in the data. Bellanoir's MaleProbability is 10, and so is Bellanoir Libero's.

The real range across all 289 breedable species is 10 to 90. Four species sit on the 10 floor and two on the 90 ceiling, and only eight species fall outside a 20-to-80 band at all — which is the other half of why the exceptions get dismissed. They are real, but most of them are mild.

Correction: it is 44 Pals with skewed ratios, not nine

Every circulating list of "the Pals with uneven gender ratios" names the same handful, usually nine. The real figure is 44 of 289. Here is all of them, grouped by value, with no rounding:

Four of those values — 38, 42, 55 and 58 — are the ones to check us on. They are the tell that a list has been read out of the game table rather than copied from another list, because nobody inventing plausible ratios invents 38% male for a single species. If a chart you are using rounds Splatterina to 40 and Whalaska to 60, it was assembled by hand.

Note also how variants travel with their base form: Dazzi and Dazzi Noct, Lyleen and Lyleen Noct, Nitemary and Nitemary Botan, Knocklem and Knocklem Ignis, Warsect and Warsect Terra, Kingpaca and Kingpaca Cryst, Elgrove and Elgrove Cryst, Bellanoir and Bellanoir Libero all carry the same value as the Pal they are a variant of. Across the whole roster exactly two variants depart from their base form's ratio, and both are worth knowing:

• Whalaska is 58 and Whalaska Ignis is 42 — the variant lands on the opposite side of even from its base form.
• Shroomer is 55 but Shroomer Noct is a flat 50, which is why Shroomer Noct does not appear in the table above at all.

So "a variant shares its base form's ratio" is a good default rather than a rule — worth checking on the species page instead of assuming.

Mozzarina was not bugged

The complaint that a Mozzarina ranch fills up with females is correct, and it has been answered wrongly for a long time. Mozzarina is 20% male, 80% female, by design, from the game's own table. Four females for every male is the intended rate, not a bad streak and not a bug.

This matters practically because Mozzarina is a Milk producer and Milk is the tightest ingredient in the Cake chain at 7 per cake. If you are breeding Mozzarina to scale up cake production, budget for the roll: you will get roughly one male in five and you only need one of them for the farm.

The six hardest gender grinds in the game

24 Pals have no cross-species breeding route at all — nothing but two of themselves produces them, so you must catch both a male and a female before you can breed one. Six of those 24 are also gender-skewed, which is the worst combination in the game: the roll is uneven and you cannot breed around it.

Everything else on the same-species-only list is 50/50, so the grind there is purely about finding the spawn. The full list of 24, with spawn counts →

Correction: passives are not more likely to come from the male parent

This is the belief worth destroying, because players restructure entire breeding chains around it — deliberately re-catching a species to get the passive onto the male, or refusing a pairing because the good parent is female. No asset we have extracted supports it.

Four independent places where a gender term would have to live, and does not:

The absence is a real absence rather than a hole in our extraction, and there is a clean way to show that. Of those ten Combi_* properties, four carry serialized values in the same asset we read — Combi_BossPalRate, Combi_PassiveInheritNum, Combi_PassiveRandomAddNum and Combi_TalentInheritNum — and six carry none. We can see values where values exist. There is no gender-weighted inheritance property in that list to have a value in the first place.

One more: search the entire shipped server binary for a gender-probability field and you get exactly one name, MaleProbability, the per-species field behind the table above. There is no second field for bred Pals, no Combi_ gender term, and nothing named for a birth or parent gender rate.

Two females, two males, and the Pal Reverser

Same-gender pairs do not breed, and the game treats that as a named state rather than a silent failure. Its breed-stopped enum has exactly six members: None, NoPair, GenderMismatch, NoItem, NoPower and SlotFull. So if your farm is idle, gender mismatch is one of six specific things it can be — and the other five are worth knowing: no pair assigned, no cake in the chest, no power, or the farm is full of uncollected eggs.

Same-gender eggs do get reported. We are not going to present that as a technique: nothing in the data describes such a routine, and with GenderMismatch shipped as a first-class stop reason we have no basis to tell you it can be bypassed.

What the game does ship is a direct fix, and it is under-documented:

That makes the Pal Reverser the real answer to "I caught four Lyleen and they are all female". You do not have to keep rolling the spawn; you have to reach level 38, build the table and find one of these.

Two things we will not tell you about it. Where to get one: the row is not craftable and not flagged for treasure boxes, and no merchant or drop table in our extraction names it. Its Price field reads 10, which is wildly out of line with comparable consumables — a plain Cake's reads 6,300 — so we are not going to present 10 as a gold cost. What its restrictions are: every effect field on the item row is zero or empty, so the behaviour is compiled into the game rather than stored in the data. Whether it flips gender or sets it, and whether anything is exempt, is not something the files answer.

Looking up one specific Pal

Every species page prints its own gender ratio next to its egg tier and breeding rank, with a callout when the roll is not even and a warning when it is both uneven and same-species-only. If you want one Pal rather than the whole list, start there — for example Mozzarina, Kingpaca or Bellanoir.

What to do next

• Plan the pairing itself with the breeding calculator — the child species is deterministic, so you can check a route before you catch anything.
• Read how the species is actually decided in the breeding guide: same species breeds true, then the override table, then the rank formula.
• Chase passives on whichever parent has them, using the passive skills guide — and Special Cake, in the cakes and Breeding Farm guide, if you want the passive count taken out of the roll.
• If a species is fighting you on gender, check whether it is on the same-species-only list before you commit to a catching grind.

Frequently asked questions