What is an ERP? Event-Related Potentials Explained Simply

Imagine that you are quietly doing your work and you suddenly hear a loud sound. Within a fraction of a second, your brain has already detected it, identified it, and reacted to it before you could fully understand what happened. This event can be seen in detail by scientists, millisecond by millisecond. To view this event unfold itself they make use of something called an Event-Related Potential, or ERP.

If you are thinking how this is possible then, here's the explanation. With the help of an EEG recording, EEG captures the continuous electrical activity of the brain and extracting ERPs from the brain's specific reaction to a single moment. In this article, I will explain what ERPs are, how they work, and why they've become one of the most powerful tools in neuroscience research.

The Difference Between EEG and ERP

EEG = It is the continuous recording of brain activity
ERP = Is the brain's specific response to a specific event or stimulus presented
In most cases the stimulus can be anything: a sound, an image, a touch, a word (depending on the need of a researcher)
The wave seen in response to the stimulus is the electrical potential change that happened in the brain.

Analogy - Think of it like EEG is recording a crowded room full of conversations and ERP is isolating what ONE person said the moment a door slammed.

Overview with the help of an example: It is like EEG acting as a security camera that records everything happening in a room 24/7. An ERP is like extracting out the exact few seconds from that footage when someone walked through the door. The camera helped to recording the whole time, but the ERP zoom helped into that one critical moment.

How Are ERPs Measured?

Most often the brains response to an event is very small and hidden under all the background noise
It is almost impossible for you to see a single trial or a trace in raw EEG.
In this case the solution is to show the same stimulus hundreds of times
Later average all the responses together.
With averaging, the random noise gets cancelled out, and the pattern that is most consistent emerges out as the brain's response.
This method is called signal averaging which is the foundation of all ERP research

Simple explanation:

Imagine that you are in a crowd trying to listen to that one person clapping inside a noisy stadium. It is impossible to that one clap because thousands of other sounds are happening at the same time.

But if you ask that one person to clap at the exact same moment 100 times, and you record the sound each time:

The random crowd noise changes with every recording
However, the clap always happens at the same time
The random noise fades away, when you average all the recordings together.
That one consistent clap becomes clear, loud and easy to recognize.

That is exactly how researchers extract ERPs from EEG: Because the brain’s response to a stimulus is small, but because it occurs at the same time on every trial, averaging reveals it and the unrelated brain activity cancels out.

What the numbers tell us?

In a typical ERP study, each stimulus is presented over 100-300 repetitions to get a true response.
The averaged results reveal peaks and valleys that occur at precise time points.
The peaks are the "components" that researchers measure

In my work every day for any kind of experiment, we present a stimulus at least 150 times to get an averaged ERP. This is not an easy process because everyday EEG typically has a lot of artefacts and to obtain clean 150 trials is a major challenge. You need these clean trails before the brain's true response reveals itself.

When I first saw a raw single-trial EEG and then the averaged the ERP, I was amazed at how a clear signal emerges from what looked like pure chaos.

Reading an ERP Waveform: Peaks, Valleys and What It Means

An ERP waveform is a graph where the time (milliseconds) is presented on the x-axis and voltage (microvolts) on the y-axis. The stimulus starts at time at zero (t=0).

After a stimulus is presented, the brain produces a series of positive and negative voltage peaks. Each peak is referred as a component.

The components are named by their polarity and timing:

P = Positive peak
N = Negative peak
The number = approximate time in milliseconds. For example: P300 = a positive peak at around 300 milliseconds

The common naming convention used:

P300 or P3 = Positive peak at ~300 ms
N100 or N1 = Negative peak at ~100 ms
P200 or P2 = Positive peak at ~200 ms
N200 or N2 = Negative peak at ~200 ms

Note:

In ERP plots, the confusing part is that the negative is traditionally plotted UPWARD.
The timing is not exactly 300 if the wave is P300, it might actually appear at 280 ms or 350 ms
The pattern is important, not the exact millisecond numbers.

It is very common for anyone new to research when looking at ERP data is panicking because their P300 appears at 340 ms instead of exactly 300ms. The naming convention is an approximation. o it is important to look at the pattern and the relative order of components, not the exact timing. The brain doesn't carry a stopwatch.

Major ERP Components and What They Tell Us

N100 (N1): The Brain's First Detection

It appears on the screen around 100 ms after stimulus
Shows the brain's initial sensory detection to "something happened"
Response seen for sounds, visual stimuli, and touch.
Larger N1 means that the brain detected the stimulus more strongly

P200 (P2): Early Classification

Appears as a response around 200ms after stimulus
This means that the brain has started classifying what it just detected
Changes by attention meaning if you are paying attention, P2 also changes.

N200 (N2): Conflict and Control

Seen around ~200 ms after stimulus is presented.
It is related to cognitive control, conflict detection, and inhibition.
In simple words it is the brain saying "wait, I need to override my automatic response"

P300 (P3): The Attention and Memory Signal

Around ~300 ms after stimulus.
This is the most studied ERP component in history.
Helps in studying - attention, surprise, and working memory updating.
The response is larger when a stimulus is rare or unexpected.
Also used in lie detection research, ADHD diagnosis, BCI.

N400 — The Language Surprise Signal

Can be seen around ~400 ms, specifically for words and language
A huge response seen when a word is unexpected or doesn't make sense in context
For example: "I drink my coffee with cream and socks" → huge N400 at "socks"

LPP (Late Positive Potential) — The Emotional Attention Signal

Is detected around 400-800 ms after stimulus.
Suggests a sustained emotional attention towards a stimulus.
A stronger response for emotional images (both positive and negative) vs neutral

In my work, I regularly measure five ERP components: N1, N2, P3, LPP, and N170. Each one tells a different chapter of the same story. When a participant sees an image, N1 tells me their brain detected it, N170 tells me if it was a face, P3 tells me if it caught their attention, and LPP tells me if it held their emotional engagement. Together, these components create a millisecond-by-millisecond timeline of how the brain processes a single moment."

Uses of ERPs in the Real World

Clinical diagnosis

ADHD - P300 is often reduced or delayed and is one of the methods used to diagnose ADHD
Schizophrenia - Abnormal mismatch negativity (MMN) and P300 seen.
Coma assessment - To check for the presence of ERPs which can indicate awareness
Hearing tests in infants - Uses auditory brainstem responses (ABR)

Cognitive research

N400 - Language processing
N170- Face recognition
P300 - Memory and attention
N200 - Decision-making and conflict

Consumer neuroscience

To test ads, packaging and product experiences
ERPs also tell us the responses that surveys cannot capture
Gives us the millisecond-level timing shows exactly WHEN engagement happens

Brain-Computer Interfaces (BCI)

P300 -based text-entry system allow paralyzed patients to type using brain signals
For instance, the system flashes letters and the brain produce P300 when the desired letter appears.
Among the most successful real-world BCI applications

References:

Why ERPs Changed How I Understand the Brain - A Research Perspective:

As a student, before hands-on experience with ERPs, I thought of brain research as looking at overall patterns, which areas are active, which aren't. Everyday working with ERPs changed that perspective.

During the analysis stage of an experiment, when I see that Stimulus A generates a larger P3 than Stimulus B, it tells me that Stimulus A it captured more attention and I also know this happened around 300 milliseconds after the person saw it, long before they could consciously evaluate it. Similarly, when I see enhanced or a bigger LPP response for one stimulus versus another, I am certain that the brain maintained emotional engagement for longer.

What makes ERPs crucial and widely used is that they capture what happens BEFORE conscious thought kicks in. However, surveys and interviews are subject to bias because it tells what people think after they've already processed and rationalized their response.

Event related potentials reveal what the brain actually did BEFORE conscious thought kicked in — the raw, unfiltered response. Hence, ERPs so valuable, and that's why I believe it is also important to understand them.

Summary

Event-Related Potentials (ERPs) are the brain's averaged electrical responses to specific events, extracted from EEG using signal averaging and a combination of different methods
Represent the brain's processing stages with millisecond precision
The major components include N1 (detection), P300 (attention), N400 (language), and LPP (emotion)
Various components are named by polarity (P/N) and approximate timing in milliseconds
ERPs are widely used for clinical diagnosis, cognitive research, consumer neuroscience, and brain-computer interfaces
They capture something that happens before conscious thought kicks in, the brain's first, unfiltered response

Frequently Asked Questions

What is the difference between EEG and ERP?

EEG is the continuous recording of all brain electrical activity. An ERP is a specific brain response to a specific event, extracted from the EEG by averaging hundreds of trials together. Think of EEG as the full movie and ERP as one freeze-frame at the exact moment something important happens.

Why is the P300 component so important?

P300 is the most widely studied ERP component because it reliably reflects attention and cognitive processing. It's used in clinical diagnosis (ADHD, schizophrenia), lie detection research, and brain-computer interfaces. Its amplitude and timing change based on how much attention the brain allocates to a stimulus.

Can ERPs be used for lie detection?

Research shows that P300 responds differently when a person recognises information they're trying to hide. However, ERP-based lie detection is not yet reliable enough for legal use and remains a research tool. The brain's response to familiar vs unfamiliar information is detectable, but individual variation makes definitive conclusions difficult.

How many trials are needed for a clean ERP?

Most researchers use 100-300 trials per condition. Fewer trials result in noisier averages. The exact number depends on the component being studied — large components like P300 may need fewer trials, while smaller components may need more.

Misconceptions About ERPs

ERPs can read specific thoughts

No, ERPs reveals different TYPES of processing (attention, emotion, conflict) and not specific content. P300 tells us that the brain noticed something unexpected that happened and not WHAT it thought about it.

You only need a few trials to get an ERP

Single-trial ERPs are mostly noise and are not considered as the most reliable response of the brain. One needs 100-300+ repetitions, and this is why ERP experiments take at least 30-60 minutes.

ERPs are outdated compared to fMRI

They give us the millisecond precision timing that the fMRI cannot match. Also helps us in understanding the TIME COURSE of brain processing. Many labs combine both the methods.

Author:

Vaishnavi Bagayi

EEG Research Associate