The Top 15 Data Center States and Where the Pipeline Is Headed
by King White, on Nov 19, 2026, 7:00:00 AM
The United States has more than 3,000 operating data centers, and more than 1,500 are in some stage of development, according to a Pew Research Center analysis of figures from Data Center Map as of Feb. 19, 2026. If that pipeline is built out, it would expand the nation’s data center base by roughly half.
The size of the pipeline is only part of the story. Where it is going matters more. A growing share of new projects is headed to states, counties and communities with little or no data center presence today. For companies deciding where to build, lease or expand data center capacity, the state-level numbers show which markets are deepening, which are leveling off and which are just getting started. They also have limits that every decision-maker should understand before drawing conclusions.
The National Picture
Planned development is concentrated in two regions. The South and Midwest account for three-quarters of all planned data centers, and the South alone holds 48% of the national pipeline.
The South has 1,209 operating data centers and 754 planned, which would increase its base by 62%. The Midwest has 655 operating and 419 planned, a 64% increase and the fastest growth rate of any region. The West has 807 operating and 277 planned, a 34% increase. The Northeast has 397 operating and 106 planned, a 26% increase.
The regional split reflects a broader change in what drives data center location decisions. Proximity to major population centers and network hubs still matters, but power availability, land and speed to market increasingly determine where the largest projects land.
The Top 15 States
The 15 states with the most data centers hold roughly three-quarters of both the nation’s operating facilities and its planned projects.
State |
Operating |
Planned |
Total |
Planned as % of operating |
| Virginia | 398 | 287 | 685 | 72% |
| Texas | 296 | 170 | 466 | 57% |
| California | 277 | 54 | 331 | 19% |
| Illinois | 139 | 123 | 262 | 88% |
| Georgia | 94 | 141 | 235 | 150% |
| Ohio | 166 | 57 | 223 | 34% |
| Arizona | 98 | 86 | 184 | 88% |
| New York | 148 | 6 | 154 | 4% |
| Oregon | 115 | 27 | 142 | 23% |
| Washington | 124 | 11 | 135 | 9% |
| Pennsylvania | 78 | 51 | 129 | 65% |
| Florida | 120 | 8 | 128 | 7% |
| North Carolina | 72 | 41 | 113 | 57% |
| Iowa | 64 | 41 | 105 | 64% |
| Indiana | 38 | 54 | 92 | 142% |
Virginia and Texas lead on every measure. Together they account for 457 planned data centers, close to 30% of the national pipeline.
Beyond the two leaders, the list mixes long-established markets such as California, New York and Washington with states where much of the activity is newer. Ohio ranks fourth in operating facilities with 166, and Pennsylvania, North Carolina and Iowa each have pipelines equal to more than half of their current base. The spread suggests the next wave of capacity will be distributed across more states than the last one.
Three Patterns in the Pipeline
The ratio of planned to operating facilities in each state reveals more than the raw totals. Three patterns stand out.
Established hubs keep growing. Virginia’s 287 planned data centers would expand its operating base of 398 by 72%, and Texas’ 170 planned projects would add 57% to its 296 operating facilities. Scale tends to attract more scale: existing fiber, experienced utilities, a trained workforce and a mature supplier network all reduce execution risk. The downside is that concentration intensifies competition for power, land and skilled labor, which puts upward pressure on costs and timelines.
In several markets, the pipeline rivals or exceeds the existing base. Georgia has 141 planned data centers compared with 94 in operation, and Indiana has 54 planned compared with 38 operating. Illinois and Arizona each have planned pipelines equal to 88% of their current base. If even a portion of these projects is completed, these states will look very different within a few years.
Some large markets have thin pipelines. New York has six planned data centers against 148 in operation, Florida has eight against 120 and Washington has 11 against 124. California’s 54 planned projects equal just 19% of its operating base. These remain significant markets, but factors such as land costs, power constraints and permitting timelines, which vary by state, appear to be steering new large-scale development elsewhere.
The Shift Toward Rural Markets
The most significant change in the data is geographic. About 67% of planned data centers are in rural areas, while 87% of operating data centers are in urban areas. In addition, 39% of planned data centers are in counties that do not have one today.
The reasons are practical. Large campuses need hundreds of acres, access to substantial power and room to expand, and rural areas can often provide all three at a lower cost than established metros. For local governments, a large project can also bring a significant new tax base.
The trade-offs are real. Rural markets often have a thinner labor pool, less existing infrastructure and utilities that may need years to build new generation and transmission. Some communities have pushed back on large projects over concerns about power costs, water use and land use, and smaller local governments may have limited experience negotiating agreements of this size. A lower-cost rural site can lose its advantage quickly if power delivery slips or local approvals stall.
What the Numbers Don’t Tell You
State-level counts are a useful indicator of market momentum, but they have real limitations, and the current pace of activity in the industry makes those limits more important. The data is a snapshot from February 2026, and projects are announced, delayed and canceled every week, so current counts in fast-moving states are likely already different.
The planned category combines projects under construction with those that are only announced or land banked, meaning a site has been secured with no firm commitment to build. Not every announced project will be completed.
Counts also treat every facility equally, so a small enterprise facility and a gigawatt-scale campus each count as one. As a result, a state with fewer facilities may carry far more capacity and power demand than its ranking suggests.
There is no federal registry of data centers, and private trackers routinely disagree on state totals because they define and count facilities differently. Finally, the counts say nothing about the factors that determine whether a location will work, including power availability and delivery timelines, incentives, workforce, water, network access and community sentiment.
What This Means for Companies Planning Data Center Capacity
For companies evaluating data center locations, state rankings are best used to understand where the market is moving, not to choose a site. A state with a large pipeline may signal strong fundamentals, but it may also signal rising competition for the same power, land and labor a company needs.
A sound location decision starts with the company’s own requirements, including power load, timeline, redundancy, latency and budget, and tests candidate markets against them. The key questions include how quickly a utility can deliver the required power, the total cost of operation including power rates and taxes, the value and durability of available incentives, and the depth of the local workforce. Established hubs offer proven infrastructure and fewer unknowns at a higher cost. Emerging and rural markets can offer land, power and incentive advantages, but they carry more execution risk.
Timing also matters. In markets with large pipelines, the best-positioned sites and early places in a utility’s interconnection queue are often claimed well before projects are announced publicly. Companies that engage utilities and economic development officials early, before their plans become public, generally have more options and more leverage on incentives. Keeping two or three viable markets in play through negotiations, rather than committing to one too soon, preserves that leverage.
This is where objective analysis matters. Site Selection Group is a tenant-only, vendor-neutral advisory firm. We do not represent landlords, developers or utilities, and we have no financial interest in which market a company chooses. That independence allows us to compare markets on the factors that matter to a company’s operation and to help it negotiate power, incentives and real estate terms from a position of strength.
To discuss how the shifting data center map affects your plans, contact Site Selection Group.
