Where are metals on the periodic table? Metals occupy roughly the left three-quarters of the table: almost all of groups 1 through 12, the lower-left corner of groups 13 through 16, and the two long rows pulled out at the bottom (the lanthanides and actinides). The dividing line runs diagonally from boron near the top toward astatine near the bottom right.
That line is not a wall. It is a gradient, and the chemistry across it explains why one element conducts electricity while its neighbor is a brittle insulator. This explainer covers the mechanism behind the divide, the counts and thresholds involved, and where the simple "left of the staircase" rule quietly fails.
What Actually Separates a Metal From a Nonmetal
The defining behavior is how an element's atoms hold their outermost electrons. Metals have loosely bound valence electrons that detach easily and drift through the solid as a shared "sea" of charge. Nonmetals hold those same electrons tightly, either in discrete atoms or in covalent bonds that keep electrons localized between two nuclei.
That one difference generates every property people associate with metals. Mobile electrons carry current, so metals conduct. They also scatter light, which is why polished metals are reflective. And they slide past each other under stress, which is why a copper wire bends instead of shattering.
Moving right across a period, each element adds one proton to the nucleus and one electron to the same outer shell. The added nuclear charge pulls that shell inward with a tighter grip, because the electrons are not screening each other effectively at the same distance. By the time you reach the right side of the table, the outer electrons are held too firmly to roam. That trend, quantified as effective nuclear charge, is the engine behind the entire metal-to-nonmetal transition.
Energy data make the gradient concrete. Ionization energy, the energy needed to remove an electron, runs from about 376 kilojoules per mole for potassium up to about 1,402 for neon, and the climb is steady rather than abrupt across a period.
Groups 1 Through 12: The Unambiguous Metal Zone
The s-block and d-block are metals without argument. Group 1 holds the alkali metals, from lithium down to francium. Group 2 holds the alkaline earths, from beryllium to radium. Groups 3 through 12 are the transition metals, and every one of them is a metal by structure and behavior.
The chemistry of these blocks maps onto their position directly. Group 1 metals carry a single valence electron, so they shed it readily and react violently with water; reactivity increases from lithium to cesium as the outermost electron sits farther from the nucleus. Group 2 metals hold two, so they are harder and less reactive but still firmly metallic.
The transition metals are where the d-orbital electrons do double duty. Because d orbitals are close in energy to s orbitals, transition metals can lose a variable number of electrons, which is why iron forms both Fe²⁺ and Fe³⁺ while sodium forms only Na⁺. That variability underlies their catalytic behavior and their wide range of oxidation states.
Two structural quirks belong here. Mercury, at the far right of the d-block, is liquid at room temperature because relativistic effects tighten its 6s orbital enough to weaken metallic bonding. And the f-block rows, pulled out below the main table, harbor the lanthanides and actinides, the elements behind most rare-earth magnets.
The p-Block Staircase: Where the Metal Map Gets Complicated
Groups 13 through 16 form the contested territory. At the upper left of this block sit metals, and at the lower right sit nonmetals, with a diagonal band of metalloids between them. The dividing line steps down from boron to astatine, giving it the nickname of the staircase.
Follow groups 13 and 14 downward and you can watch the transition happen. Group 13 runs boron (a metalloid), aluminum (a metal), gallium, indium, and thallium, all metals below boron. Group 14 runs carbon (a nonmetal), silicon and germanium (metalloids), then tin and lead (metals). Element 114, flerovium, sits below lead but exists only as fleeting synthetic atoms, so its bulk metallic character is inferred rather than observed.
The metalloids themselves are the boundary case. Boron, silicon, germanium, arsenic, antimony, and tellurium are the six conventionally listed, with astatine and occasionally polonium included by some authors. The ambiguity is not sloppiness; it reflects a real continuum. Metalloids are semiconductors precisely because their valence electrons are neither fully free nor fully locked.
| Zone | Typical elements | Electron behavior | Practical signature |
|---|---|---|---|
| s-block metals | Sodium, magnesium | One or two loosely held valence electrons | Soft to hard, highly reactive |
| d-block metals | Iron, copper, gold | Variable d-orbital loss, delocalized sea | Conductive, catalytic, alloyable |
| f-block metals | Neodymium, uranium | Inner-shell electrons, high density | Strong magnets, nuclear fuel |
| Metalloids | Silicon, germanium | Partly bound, partly mobile | Semiconductors, brittle |
| p-block metals | Aluminum, lead | Few valence electrons, moderate pull | Lightweight or dense structural use |
| Nonmetals | Oxygen, chlorine | Tightly held electrons | Gases, insulators, reactive |
Counting the Metals: How Many and How Pure
Exact counts depend on definition, and the range is wider than most summaries admit. About 91 of the 118 confirmed elements show metallic character under ordinary conditions, by the broadest count that treats all transition metals, lanthanides, and actinides as metals. If you restrict to elements that are solid metals at room temperature and exclude the metalloids and synthetics, the figure drops toward the high 80s.
The number also depends on conditions. Hydrogen is a nonmetal gas at standard pressure but becomes a metallic solid under compression above roughly 495 gigapascals, an experimental threshold reported for hydrogen in 2016 (Science). That example shows why "metal" is a behavior, not a fixed label.
Relative abundance matters more than count for practical purposes. In Earth's crust, oxygen and silicon dominate at 46.6 percent and 27.7 percent by mass, and the most abundant actual metal is aluminum at about 8.1 percent, followed by iron at 5.0 percent (CRC Handbook of Chemistry and Physics). That ordering is why aluminum is cheap despite being a metalloid neighbor, and why iron anchors industrial civilization.
The counterintuitive detail runs the other direction: precious metals like gold, platinum, and rhodium are metals in the strictest sense, yet their electrons are held tightly enough by relativistic and d-orbital effects that they resist oxidation and corrosion. Being a metal does not guarantee reactivity.
Where the Simple Rule Fails
Several elements break the tidy left-versus-right picture, and each failure teaches something about the underlying mechanism.
Hydrogen sits at the top left, above lithium, but it is not a metal at ambient conditions. It holds its single electron too tightly, and its nonmetal placement reflects bonding behavior rather than geography.
Tin and lead are p-block metals that behave oddly: tin has two solid allotropes, and its common "tin pest" transformation below about 13 degrees Celsius degrades the metallic form into a powdery nonmetallic allotrope. That is a metal crossing its own category line with temperature.
Carbon can also masquerade as metallic. Graphite conducts electricity along its layers because of delocalized electrons, so a nonmetal acts partly metallic depending on how its atoms bond. The lesson is that the periodic table labels elements, but allotrope and pressure can redefine behavior.
Some textbooks still shade astatine as a metalloid and others as a halogen nonmetal. Its scarcity, with a half-life of about 8.1 hours for its longest-lived isotope, means its behavior is inferred more than measured.
Red-flag warnings for anyone reading a metal-boundary claim:
A chart that shades mercury as nonmetallic is wrong; it is a d-block metal that happens to be liquid. A source listing exactly one count of metals without stating its definition is unreliable, since the count varies by methodology and conditions. Claims about the "metalloid staircase" that omit astatine's disputed status are oversimplifying. Statements that a specific synthetic element is "a metal" without noting they were never observed in bulk overstate certainty.
Using the Map in Practice
If you need to locate or reason about an element's metal status, a few concrete steps get you there reliably.
Start with the block, not the landmark. Groups 1 through 12 are metals; groups 17 and 18 are nonmetals. Spend your doubt only on groups 13 through 16. Apply the staircase to the p-block. Elements above and to the right of the boron-to-astatine line are nonmetals or metalloids; below and left are metals. Check ionization energy against the trend. Values near or above about 1,000 kilojoules per mole usually signal a nonmetal or metalloid, while values under about 600 typically mark a metal. Consult the density and conductivity data. Bulk metals generally exceed 2 grams per cubic centimeter as solids and support a measurable current; metalloids fall in intermediate ranges. Note the row. For borderline elements, metallic character increases downward, so the same group can hold a metalloid at the top and a metal well below it.
These thresholds reconcile the counted range of 88 to 91. The narrower numbers exclude synthetics and borderline p-block entries; the broader numbers include them. If you state a figure, state the definition that produced it, because a bare count invites a contradiction you did not intend.
What the Map Cannot Tell You
The periodic table predicts a great deal about metal status, but it sets limits that prevent overreach. It tells you class, not detail: it cannot predict alloy strength, corrosion rates, or which oxidation state a transition metal will adopt in a given compound. Those require experimental data and quantum modeling.
Position also fails to settle liquid and solid phases under extreme conditions. Pressure and temperature can metallicize substances the table places as nonmetals, and new synthetic elements push the count in either direction. Ongoing research on superheavy elements continues to revise where the metallic frontier lies.
Practical use of the table should therefore treat its metal boundary as a reliable guide rather than a rule of physics. The staircase is a map, and all maps generalize. For engineering decisions, the element's measured properties, not its address, decide the matter.
The same care applies to interpreting the metalloid zone. It is best understood as a transition band where the physics shifts continuously, which is why silicon powers electronics while its neighbors in the same period are inert or metallic. Read the table with that gradient in mind, and the placement of metals stops being a memorization exercise and becomes the explanation.