Croatia
Croatia iGaming market in numbers
| Metric | 2025 | 2026 |
|---|---|---|
| Total GGR | $370m | $400m |
| Regulated GGR | $250m | - |
| Offshore GGR | $120m | - |
| Channelization | 68% | - |
| Mobile share | 75% | - |
| YoY growth | - | +8.0% |
| CAGR 2021–2026 | +9% | - |
Regulated and offshore split
Legal status by vertical
Operator's read on Croatia
Croatia is a regulated European market that is, in practice, closed to standalone foreign online operators, and that single fact should frame the entry decision. The Ministry of Finance regulates gambling, and online licences are effectively restricted to operators that also hold a Croatian land-based concession and maintain local presence, including a Croatian office and servers. The strategic point is that Croatia is not a market you apply into from abroad. It is a market you enter through local partnership or acquisition, much as the larger operators have done.
The local-establishment requirement is the gating fact. Because an online licence is tied to a land-based concession and a genuine local presence, a foreign operator cannot obtain a pure-online Croatian licence on its own. The realistic route in is to partner with or acquire a licensed local operator, following the model of SuperSport, which Entain acquired in 2022. That makes Croatia a corporate-development decision as much as a market-entry one, and an operator without a local vehicle does not have a route in.
The 2025-26 reform is raising costs and compliance load. A new Gambling Act reform is being phased in through 2025 to 2026, with player-protection measures already rolling out: a national self-exclusion register, mandatory identification, an advertising ban from 06:00 to 23:00 on television and radio, and venue-distance rules, with the full law targeted for 2026. The reform tightens the operating environment and signals a regulator focused on harm reduction, so an entrant should expect a stricter regime rather than a stable status quo.
The licence costs are rising sharply. Licence fees are climbing under the reform, with the online gambling licence moving from around €265,445 toward €398,168, alongside a significant share-capital requirement and a 15-year term. Those numbers raise the bar for entry and reinforce that Croatia is a serious commitment rather than a light-touch market. Combined with the local-establishment requirement, the rising fees mean only operators prepared to invest in a genuine Croatian presence should consider it.
What winning looks like. Winning in Croatia looks like a local partnership or acquisition that satisfies the concession and establishment requirements, an operation built for the tightening player-protection regime, and a brand localised for the Croatian player. The operators who succeed treat the local vehicle as the prerequisite it is and build the rest of the proposition around it, rather than hoping to serve the market from a foreign base.
The regional play. Croatia sits among the Balkan and central European markets, and it suits operators willing to acquire or partner locally as part of a broader regional footprint rather than those seeking a standalone foreign licence. How a partnership-gated market like Croatia fits an entry plan is part of the multi-market sequencing piece.
The biggest mistake. The biggest mistake is treating Croatia as an open market you can license into from abroad, when an online licence requires a local concession and establishment. The related mistake is underestimating the rising fees and the tightening reform. Plan a local partnership or acquisition first, build for the stricter regime, and only commit once the local vehicle and the economics are in place.
What's changing
Stable framework.
Where these figures come from
- Croatian Tax Administration 2024
GGR figures are 2025 estimates or actuals where regulator data is available; 2026 projections drawn from the most recent published forecasts. Offshore figures are inherently more uncertain than regulated figures and should be treated as directional. Where reputable sources disagree materially the dataset uses the midpoint of the range.