Exponential Distribution Calculator

Exponential Distribution Calculator

Find probabilities, quantiles, the mean, median and variance of an exponential distribution — and see the memoryless property in action. Enter the rate λ or the mean: λ = 1 / mean.

Events per unit time
Average waiting time
Time to check (x ≥ 0)
P(X < 1) · λ = 0.5
0.3935

About 39.35% of waiting times are shorter than 1 unit — the mean wait is 2, so most waits are short and a few are very long.

Mean
2.0000
Median
1.3863
Variance
4.0000
Standard deviation
2.0000
P(X > 1)
0.6065
90th percentile
4.6052

Calculation

    Density curve — the region you asked about

    The shaded area under the PDF is P(X < 1) = 0.3935. Switch the curve to CDF or survival below.

    Memoryless: P(X > 5 | X > 3) = P(X > 2) = 0.3679 — a bulb that has already lasted 3 hours is as good as new.
    Try the memoryless property yourself

    P(X > 5 | X > 3) = 0.3679 and P(X > 2) = 0.3679 — they are always equal, whatever s and t you pick.

    An exponential distribution calculator finds probabilities, quantiles and the mean, median and variance of a waiting time — the time until the next event in a Poisson process. Enter the rate λ (events per unit time) or the mean (1 / λ) and it returns P(X < x), P(X > x), P(a < X < b) or the quantile for any probability, with the density curve and a memoryless-property check.

    How to use the exponential distribution calculator

    1. Enter the rate λ — or the mean. The calculator keeps the two in step: λ = 1 / mean.
    2. Choose what to find: P(X < x), P(X > x) or P(a < X < b). Or switch to Quantile mode and enter a probability p to find the value x with P(X < x) = p.
    3. Read the headline result, then the results grid: mean, median, variance, standard deviation, survival and the 90th percentile. The calculation ledger shows every step.
    4. Check the curve: the shaded area is the probability you asked for. Switch the curve to CDF or survival, and try the memoryless demo below the chart.

    Try it now: with λ = 0.5 (mean 2), P(X < 1) = 0.3935, P(X > 3) = 0.2231 and P(1 < X < 3) = 0.3834. The median is 1.3863 and the 90th percentile is 4.6052.

    What is the exponential distribution?

    The exponential distribution models the time between events in a Poisson process — the wait until the next bus, the next phone call, the next radioactive decay. It has a single parameter, λ, the rate. Its density is f(x) = λe−λx for x ≥ 0, its CDF is F(x) = 1 − e−λx and its survival function is S(x) = e−λx. The mean is 1 / λ, the median is ln 2 / λ, the variance is 1 / λ² and the standard deviation is 1 / λ. It is one of the types of probability distributions you will meet first, and the continuous partner of the Poisson distribution calculator.

    λ is a rate, the mean is a time. If buses arrive at λ = 0.5 per hour, the mean wait is 1 / λ = 2 hours. A bigger λ means events come faster and the mean wait is shorter: λ = 2 per hour means a mean wait of 0.5 hours. Never mix the two up — it is the number-one student error with this distribution.

    The memoryless property

    The exponential distribution is memoryless: the past does not matter. The chance of waiting t more units, given you have already waited s units, is the same as the chance of waiting t units from the start: P(X > s + t | X > s) = P(X > t). A light bulb that has already burned for 3 hours is, in this model, as good as a new one: with λ = 0.5, P(X > 5 | X > 3) = e−2.5 / e−1.5 = 0.3679, exactly the same as P(X > 2) = e−1 = 0.3679. Try it in the demo below the chart — pick any s and t and the two numbers always agree. The exponential is the only continuous distribution with this property; its discrete cousin, the geometric distribution, is the only discrete one.

    The Poisson-process link

    If events arrive as a Poisson process with rate λ, the waiting time between events is Exponential(λ). That is why the two distributions share a parameter: the Poisson distribution counts how many events happen in a fixed time, and the exponential distribution measures how long you wait for the next one. P(no event in 3 hours) = e−λ · 3 = e−1.5 = 0.2231 — the same number as P(X > 3), because “no event in 3 hours” and “the wait exceeds 3 hours” are the same event. Read more in the types of probability distributions guide.

    QuantityFormulaλ = 0.5
    PDF f(x)λe−λx0.5e−0.5x
    CDF F(x) = P(X ≤ x)1 − e−λx1 − e−0.5x
    Survival S(x) = P(X > x)e−λxe−0.5x
    Mean1 / λ2
    Medianln 2 / λ1.3863
    Variance1 / λ²4
    Standard deviation1 / λ2
    90th percentile−ln(0.1) / λ4.6052

    Worked examples

    Example 1 — probabilities, median and quantiles (λ = 0.5)

    Waiting times follow an exponential distribution with rate λ = 0.5 per hour, so the mean wait is 1 / 0.5 = 2 hours. P(X < 1) = 1 − e−0.5 = 1 − 0.6065 = 0.3935, so about 39.35% of waits are under an hour. P(X > 3) = e−1.5 = 0.2231, so about 22.31% of waits exceed 3 hours. P(1 < X < 3) = 0.6065 − 0.2231 = 0.3834. The median is ln 2 / 0.5 = 1.3863 hours — half of all waits are shorter than this, even though the mean is 2, because the distribution is right-skewed. The variance is 1 / 0.5² = 4 and the standard deviation is 2. The 90th percentile is −ln(0.1) / 0.5 = 4.6052 hours. And the memoryless check: P(X > 5 | X > 3) = 0.3679 = P(X > 2).

    Example 2 — the Poisson-process bridge (λ = 0.5 per hour)

    Calls arrive at a help desk as a Poisson process at λ = 0.5 per hour. The waiting time until the next call is Exponential(0.5). What is the chance of no call in 3 hours? P(no event in 3 hours) = e−0.5 · 3 = e−1.5 = 0.2231. And P(the wait exceeds 3 hours) = e−1.5 = 0.2231 — the same number, by construction. Check it against the Poisson distribution calculator: P(0 events in 3 hours) with mean λt = 1.5 is also 0.2231.

    Edge cases

    With λ = 2 (mean 0.5), P(X < 0.25) = 1 − e−0.5 = 0.3935 — the same 0.3935 as Example 1, because λx = 0.5 in both cases; only the product λx matters. With λ = 1, P(X > 1) = e−1 = 0.3679. At x = 0, P(X < 0) = 0 — a wait cannot be negative, so the calculator asks for x ≥ 0. A very large λ piles the whole distribution up near zero: with λ = 10 the mean wait is 0.1 and P(X > 1) = e−10 = 0.0000 to four decimal places. Unlike the normal distribution, which is symmetric, the exponential is always right-skewed — the mean (2) sits above the median (1.3863) for λ = 0.5.

    Frequently asked questions

    What does λ mean?

    λ (lambda) is the rate: how many events happen per unit of time, on average. If λ = 0.5 per hour, you expect one event every 2 hours — the mean waiting time is 1 / λ = 2 hours. A bigger λ means events come faster and the mean wait is shorter: λ = 2 per hour means a mean wait of 0.5 hours. λ is a rate, not a time — never confuse the two.

    What is the memoryless property?

    The memoryless property says the past does not matter: the chance of waiting t more units, given you have already waited s units, is the same as the chance of waiting t units from the start — P(X > s + t | X > s) = P(X > t). With λ = 0.5, P(X > 5 | X > 3) = 0.3679, exactly the same as P(X > 2). A light bulb that has already burned for 3 hours is, in this model, as good as a new one. The exponential is the only continuous distribution with this property.

    Source: the exponential distribution on this page is the standard one-parameter distribution with PDF f(x) = λe−λx. All worked examples on this page were recomputed with Python (scipy.stats). Last updated 19 November 2026.

    See all 91 free statistics tools on Empirical Rule Calculators.