Pick one or many calendar dates from an exact eligible pool — with weekday filters, uniqueness, exclusions, formats and reproducible seeded draws when you need them.
Generate random dates
Define the eligible calendar days first, then draw uniformly from that pool.
First random result
—
Eligible: —Browser cryptoUnique draw
#
Date
Weekday
ISO value
Generate a batch to see results.
More options
UTC avoids daylight-saving shifts changing the calendar date.
Include these weekdays
Randomness: blank seed uses crypto.getRandomValues() with rejection sampling so eligible dates are not biased by modulo arithmetic. A seed intentionally switches to deterministic pseudo-random output for repeatable test fixtures. Unique mode samples without replacement. This is suitable for ordinary sampling/test data, not a certified regulated drawing.
Core random, wheel & group workflows
Use the broad picker, number, wheel, name, team or dice workflow first. Move to a specialist only when the probability model or output structure genuinely changes.
Random selection can make equal-probability choices from the entered pool, but it cannot prove that the input list itself is appropriate or unbiased. Visual wheel/dice animations are presentation only. Seeded tools remain reproducible pseudo-random workflows rather than cryptographic randomness, and rating-balanced teams use a transparent heuristic rather than an optimal or human-judgment model.
Practical guide and verification
Use the product workflow above first. These notes add interpretation, verification and limits below the interactive result area.
Define the eligible pool before drawing
The start and end dates, weekday restrictions and uniqueness setting determine which calendar days can be selected. Check that the eligible pool is at least as large as the requested unique count before interpreting a failed draw as a randomness problem.
Weekday-only sampling changes the population
Excluding weekends does not merely hide Saturday and Sunday after generation; it changes the set of eligible outcomes. That is appropriate for workday test data, but it should not be used when every calendar day is supposed to have equal opportunity.
Unique draws are sampling without replacement
When uniqueness is enabled, selecting one date removes it from the remaining pool. The dates are therefore not independent draws with replacement. This distinction matters when the output is being used for simulations or statistical sampling.
Use ISO dates for machine-readable test data
YYYY-MM-DD is unambiguous and sorts chronologically as plain text. Locale-style numeric dates can be misread across regions. Keep the underlying date values stable and change only the display format when the destination has a presentation requirement.
Random dates are not a certified drawing record
Browser randomness is useful for test data, classroom choices and ordinary sampling. It does not by itself create an independently reviewable or legally certified selection process. For consequential draws, preserve the required external review procedure instead.