Open comparison →
STRUCTURED COMPARISON
Build a comparison
Choose two categories from the same family. The same properties, equal visual weight, sources next to the claims they support.
STRUCTURED COMPARISON
Choose two categories from the same family. The same properties, equal visual weight, sources next to the claims they support.
Technology
Incandescent bulbs produce light by thermal radiation. Electric current meets resistance in a thin tungsten filament, heating it to roughly 2,500-3,000 kelvin until it glows and emits visible light along with substantial infrared. The filament sits in a glass envelope filled with inert gas or held under vacuum to slow evaporation. Because emission follows the physics of a hot body, the light is a smooth continuous spectrum set almost entirely by temperature. This mechanism is simple and well understood, but it is inherently tied to heating the filament, which shapes efficiency, colour and lifespan discussed in the rows that follow.
LED bulbs create light through electroluminescence: an electrical current passes through a semiconductor diode, and electrons releasing energy emit photons directly, without heating a filament. A small driver circuit converts mains voltage to the low-voltage direct current the chip requires. Because emission comes from the material's electronic structure, the spectrum depends on the semiconductor and any phosphor coating used to produce white light. Public technical bodies describe this solid-state mechanism as fundamentally different from thermal sources. Actual behaviour varies with chip quality, driver design and heat management, so two LED products using the same basic principle can perform quite differently.
Incandescent bulbs are generally characterised by energy agencies as low-efficiency light sources, typically delivering only around 10 to 15 lumens per watt. The large majority of the electrical energy becomes heat rather than visible light, a direct consequence of the thermal-radiation mechanism. Because of this low efficacy, several jurisdictions have restricted or phased out general-service incandescent lamps through efficiency regulations. Exact figures depend on wattage and design, and specialised or rough-service types differ. The physics that gives incandescent light its continuous spectrum is the same physics that limits its efficiency, so improvements within the technology are inherently constrained.
Incandescent bulbs release most of their input energy as heat, both as infrared radiation and as conduction through the glass, since heating the filament is how they make light. Surface temperatures can become high enough to pose burn or fire risks near flammable materials, and this is reflected in safety guidance. The same waste heat has occasionally been used deliberately, for example in small heat lamps or older enclosed appliances. In normal lighting use, however, this heat represents the energy that does not become visible light. The strong thermal output is a direct, unavoidable feature of the incandescent mechanism rather than a design flaw that can be tuned away.
Light affects circadian rhythms and sleep, and through them well-being and performance. The consensus lighting recommendations are formulated not in the names of technologies but in a measurable quantity — melanopic equivalent daylight illuminance under an international standard. The question is therefore settled by the spectrum and level of light at the eye, not by whether the lamp is incandescent.
Exactly the same rule applies to LED light: the recommendations are expressed in melanopic equivalent daylight illuminance, not in the type of source. The spread between models in colour temperature and colour rendering is wide, so nothing about the effect on circadian rhythms can be concluded from the name of the technology — the characteristics of the particular lamp are needed.
Incandescent bulbs typically have the lowest purchase price, which can make them appear cheaper at the point of sale. Over time, however, their low efficiency raises electricity use and their short rated life means more frequent replacement, both of which add to the total cost of ownership. How large that running cost becomes depends heavily on usage hours, local electricity prices and how the lamp is operated, so comparisons vary between households and countries. Some specialised or infrequently used applications may see little difference. Because the outcome is so context-dependent, reliable universal cost figures are limited, and estimates should be read as scenario-based.
The incandescent lamp is a rare case of a category leaving the market not by the buyer's choice but by a regulator's decision. A study traces this phase-out in Europe after the ban entered into force, from 2000 to 2020, and examines the processes that accompanied it. This concerns the European market: elsewhere the timing and scope of such measures differ.
LED lamps took the place of a category withdrawn from the European market by regulation. The requirement itself did not go away: it applies to the efficiency of a light source rather than to its construction, and it covers the replacement products in the same way. The study tracing the incandescent phase-out describes exactly this transition.
Counted for this table only, not for the whole site.
Both lamps produce light from electricity, but through different physics: thermal radiation from a glowing filament against electroluminescence in a semiconductor. The differences run along energy efficiency, rated life and heat output. Assessments of light quality disagree: the filament gives a continuous spectrum, while an LED's colour rendering and flicker depend on the model, so the comparison runs into the spread between products. Total cost of ownership depends on tariffs and hours of use, so there is no single figure for it.
FREQUENTLY EXPLORED
LED lighting is generally described by energy agencies as among the most efficient widely available lighting, commonly delivering on the order of 80 to over 100 lumens per watt at the product level. Most of the input energy is converted to visible light rather than heat, though driver losses and thermal conditions reduce real-world figures. Reported efficacy varies considerably between products, brands and operating temperatures, and laboratory ratings may exceed everyday performance. Independent standards and testing programmes exist to verify claims. Efficiency also tends to improve across product generations, so newer models often deliver more lumens per watt than older ones under comparable conditions.
Incandescent bulbs are typically rated at around 1,000 hours of service under standard test conditions, though values vary by design and voltage. Life is limited by gradual evaporation and eventual failure of the tungsten filament, which usually breaks suddenly rather than dimming slowly. Running a lamp at higher voltage increases brightness and efficacy but shortens life, while lower voltage extends it; long-life variants trade some light output for durability. Because the wear mechanism is physical filament degradation, ratings are comparatively predictable but short. This relatively brief lifespan is one reason efficiency policies and total-cost comparisons weigh replacement frequency so heavily.
LED products are commonly rated for long service lives, with manufacturer and programme figures often in the range of roughly 15,000 to 50,000 hours, and some higher. Rather than failing abruptly, LEDs typically dim gradually, so rated life is often defined as the point where output falls to a set percentage of the initial level. Real longevity depends heavily on heat management, driver reliability and duty cycle; poor thermal design can shorten life well below the rating. Testing standards and voluntary labelling programmes aim to make lifetime claims comparable. Because failure modes differ from older lamps, rated hours should be read alongside these conditions.
Incandescent bulbs emit a smooth, continuous spectrum characteristic of a heated body, giving a warm colour with a correlated colour temperature typically around 2,700 kelvin and a colour rendering index close to the maximum. Because colour is set by filament temperature, it is essentially fixed for a given lamp and shifts warmer as the bulb is dimmed, behaviour many users find pleasant. There is generally no visible flicker at mains frequency and no phosphor gaps in the spectrum. The trade-off is that this high colour quality is inseparable from the inefficient, hot mechanism producing it, so it cannot be adjusted like electronic sources.
LED bulbs are sold across a wide range of correlated colour temperatures, from warm to cool white, and increasingly in tunable and dimmable forms. Colour rendering, measured by indices such as CRI, varies between products; many general-service LEDs render colours well, but cheaper units can score lower. Some LEDs also exhibit flicker or produce a spectrum with gaps compared with a continuous thermal source, which is why measurement bodies work on standardised metrics. Perceived quality therefore depends strongly on the specific product and its driver, and evidence on comfort effects is mixed, making this an area where product-level testing matters.
LED bulbs convert a larger share of energy to light and therefore radiate less heat forward than incandescent lamps, and they emit little infrared in the beam. However, they are not heat-free: the chip and driver generate waste heat that must be carried away by a heatsink or the housing, which is why many LED bulbs have metal or finned bases. Excessive junction temperature is a leading cause of reduced output and shortened life, so thermal design is central to performance. Enclosed fixtures can trap heat and affect rated bulbs. Overall surface and emitted heat are generally lower, but management remains an engineering concern.
LED bulbs usually carry a higher purchase price than incandescent lamps, though prices have fallen substantially over time. Their running cost is lower because of higher efficiency and longer rated life, which reduces both electricity use and replacement frequency. Energy agencies present lifetime cost estimates suggesting savings in many typical households, but the actual payback depends on how many hours the bulb runs, local electricity tariffs, product quality and whether the rated life is achieved. Because these variables differ widely by region and usage, a single universal figure is not reliable, and published estimates should be treated as illustrative rather than guaranteed.
Evidence reviewed: September 3, 2026
Sources are reviewed by the CONTRAST editorial team. External links open in a new tab.