⚙️ How-To Dams & reservoirs Notebook By · Updated

How Old Are America's Dams? Completion Years by Decade and State

On this page

Problem statement

Most American dams were built in a burst after the Second World War: farm ponds and small flood-control dams by the tens of thousands, many of them funded by federal soil-conservation programmes, alongside the large federal dams of the West. The National Inventory of Dams records the year each dam was completed, so the age of the fleet can be counted by decade and by state. A fifth of the dams have no year, and the years that are recorded cluster on round numbers, so the counts need care at both ends. This page measures how old the dams are and when they were built.

Quick answer

Take yearCompleted above 0, count by decade, and compute ages from 2026:

y = nid.yearCompleted.where(nid.yearCompleted > 0)
decade = (y // 10 * 10).astype("Int64")
print(decade.value_counts().sort_index().loc[1900:].to_string())
print(f"with a year: {y.notna().sum():,} of {len(nid):,}; median age in 2026: {2026 - y.median():.0f} years; built before 1970: {(y < 1970).mean():.1%}")
yearCompleted
1900     1800
1910     1891
1920     2134
1930     3745
1940     4018
1950    11119
1960    18622
1970    12842
1980     5454
1990     4665
2000     3706
2010     1077
2020      243
with a year: 73,708 of 92,626; median age in 2026: 61 years; built before 1970: 49.4%

73,708 dams have a completion year, and the median one was finished in 1965, 61 years ago. The 1960s alone built 18,622 of them, and the 1950s and 1970s another 24,000 between them. Since 2010 only 1,320 dams with a year have been added. 49.4% of the dams with a year were built before 1970.

Bars of dams completed per decade from 1850 to 2020, rising to a peak of 18,622 in the 1960s and falling steeply after the 1970s. Bars of dams completed per decade from 1850 to 2020, rising to a peak of 18,622 in the 1960s and falling steeply after the 1970s.
Dams by decade of completion.

Counting by decade

1. Dams without a year

print(nid[y.isna()].publicHazardId.value_counts(normalize=True).round(3).to_string())
print(nid[y.notna()].publicHazardId.value_counts(normalize=True).round(3).to_string())
publicHazardId
Low             0.545
Undetermined    0.184
Significant     0.137
High            0.133
publicHazardId
Low             0.678
High            0.196
Significant     0.119
Undetermined    0.008

The dams without a year are not a random sample. 18.4% of them have an undetermined hazard class, against 0.8% of dams with a year, and only 13.3% are high hazard against 19.6%. They are mostly small, little-documented private dams, so the age figures describe the better-recorded part of the fleet.

2. The 1960s peak by purpose

t = pd.crosstab(decade, nid.primaryPurposeId).loc[1930:2020]
print(t[["Recreation", "Flood Risk Reduction", "Fire Protection, Stock, Or Small Fish Pond", "Irrigation", "Hydroelectric"]].to_string())
primaryPurposeId  Recreation  Flood Risk Reduction  Fire Protection, Stock, Or Small Fish Pond  Irrigation  Hydroelectric
yearCompleted                                                                                                            
1930                    1546                   140                                         534         395            162
1940                    1386                   238                                         833         491            119
1950                    3651                  1516                                        2213        1228            168
1960                    5117                  4838                                        2827        1199            166
1970                    3292                  2911                                        1900        1101             69
1980                    1012                  1571                                         562         443            105
1990                     740                  1262                                         659         214             40
2000                     569                  1098                                         400         229              4
2010                     165                   280                                          37          98              0
2020                      20                    48                                           1          13              1

The 1960s peak is broad, but its leaders are recreation dams (5,117) and flood-control dams (4,838). Stock and fish ponds peaked at the same time, while hydroelectric dams were built at a steady 160 or so a decade from the 1930s to the 1960s and then almost stopped. Since 2000, flood risk reduction has been the main reason dams are built.

3. Round years

print(y.mod(10).value_counts().sort_index().astype(int).to_string())
yearCompleted
0.0    13102
1.0     5729
2.0     6334
3.0     6118
4.0     6691
5.0     8771
6.0     6629
7.0     6713
8.0     7192
9.0     6429

Years ending in 0 are twice as common as any other digit, 13,102 against 5,700 to 7,200, and years ending in 5 hold 8,771. Many years are estimates rounded to five or ten. That is why this page counts by decade: a dam estimated at 1960 lands in the 1960s whether it was built in 1958 or 1962, so the decade bars are much less affected than single years.

Lines of dams completed per decade by purpose from 1930 to 2020, with recreation and flood risk reduction both peaking near 5,000 in the 1960s. Lines of dams completed per decade by purpose from 1930 to 2020, with recreation and flood risk reduction both peaking near 5,000 in the 1960s.
Dams built per decade by primary purpose.

Age by state and by size

The data used on this page

The US Army Corps of Engineers' National Inventory of Dams (NID), the whole inventory as a GeoPackage from its API, downloaded on 19 October 2026 and saved on this site, because the inventory is updated as states and agencies send revisions. Heights are in feet, storage in acre-feet and areas in acres, as the NID gives them.

from pathlib import Path
import zipfile

import geopandas as gpd
import numpy as np
import pandas as pd
import requests

Path("data").mkdir(exist_ok=True)
# US Army Corps of Engineers, National Inventory of Dams: the whole inventory as a GeoPackage, and the Census
# Bureau's 2024 cartographic boundary counties. Saved 19 October 2026, because the inventory is updated continuously.
# Live: https://nid.sec.usace.army.mil/api/nation/gpkg
ARCHIVE = "usace-nid-2026-10-19.zip"
if not Path(f"data/{ARCHIVE}").exists():
    Path(f"data/{ARCHIVE}").write_bytes(requests.get(f"https://data.spatialworkflow.io/{ARCHIVE}", timeout=900).content)
Z = zipfile.ZipFile(f"data/{ARCHIVE}")
if not Path("data/nid.gpkg").exists():
    Z.extract("nid.gpkg", "data")
NUMERIC = ["damHeight", "hydraulicHeight", "structuralHeight", "nidHeight", "damLength", "volume", "yearCompleted",
           "nidStorage", "maxStorage", "normalStorage", "surfaceArea", "drainageArea", "maxDischarge"]
nid = gpd.read_file("data/nid.gpkg")
nid[NUMERIC] = nid[NUMERIC].apply(pd.to_numeric, errors="coerce")       # the GeoPackage stores them as text
print(f"National Inventory of Dams: {len(nid):,} rows, {nid.nidId.nunique():,} NID ids")
National Inventory of Dams: 92,626 rows, 91,840 NID ids

Example 1: the oldest and youngest states

age = (2026 - y).groupby(nid.state).median()
n = y.groupby(nid.state).count()
age = age[n >= 200].sort_values()
print(age.head(4).to_string(), "\n...", age.tail(4).to_string(), sep="\n")
state
Nevada      36.0
Iowa        48.0
Missouri    52.0
Maryland    53.0

...
state
New Hampshire     96.0
Massachusetts    109.0
Maine            111.0
Connecticut      123.0

Among states with at least 200 dated dams, Nevada's are the youngest, with a median age of 36 years, then Iowa, Missouri and Maryland. New England's are the oldest by far: Connecticut's median dam is 123 years old, Maine's 111 and Massachusetts's 109, so half of Connecticut's dated dams were finished before 1904.

Horizontal bars of median dam age by state: youngest Nevada 36, Iowa 48 and Missouri 52 years; oldest Connecticut 123, Maine 111 and Massachusetts 109. Horizontal bars of median dam age by state: youngest Nevada 36, Iowa 48 and Missouri 52 years; oldest Connecticut 123, Maine 111 and Massachusetts 109.
The youngest and oldest dams by state.

Example 2: big dams and small dams

print((2026 - y).groupby(nid.nidHeightId).median().to_string())
nidHeightId
25-50 feet               59.0
51-100 feet              59.0
Greater than 100 feet    64.0
Less than 25 feet        66.0
Undetermined             32.0

Size makes little difference: the median dam under 25 feet is 66 years old, and those of 25-50 and 51-100 feet are 59. Dams over 100 feet have a median age of 64, from the era of the great federal projects.