Pine ScriptRSI settings

Best RSI settings for 1-minute charts

There is no universal best setting. Start by testing RSI 7 with 80/20 levels against the TradingView default, RSI 14 with 70/30 levels. Keep the same symbol, session, fees, and date range so the comparison is fair.

A useful starting point, not a trading promise

On a 1-minute NQ or QQQ chart, RSI 7 reacts faster than RSI 14. Wider 80/20 bands reduce the number of extreme readings. You still need a backtest because speed and signal quality are not the same thing.

An RSI reading describes momentum. Pineify turns the rule you choose into executable TradingView Pine Script so you can test the rule before using it.

Settings table

Compare RSI 5, 7, 9, and 14

These are test candidates. The table describes how the lookback changes the indicator, not which setting will make money.

RSI settings to test on 1-minute charts
RSI lengthLevelsBehaviorTest use
580 / 20Fastest response and the most sensitivity to single-bar moves.A stress-test candidate for very short holding periods.
780 / 20Faster than the default while keeping a short smoothing window.A practical first test for 1-minute NQ or QQQ charts.
975 / 25Slower than RSI 7, with fewer threshold crosses.A middle option when RSI 7 reacts too often.
1470 / 30TradingView default and the slowest baseline in this comparison.The control setting for every short-period test.

Keep the market fixed

Test NQ and QQQ separately. Their sessions and price series do not create identical one-minute bars.

Keep costs in the model

Commission and slippage matter more when a rule trades frequently and targets small moves.

Keep a holdout period

Pick settings on one date range, then test them on dates you did not use during selection.

Pine Script v6

Copy an RSI settings tester

Change the length and levels in TradingView without editing the formula. The script marks crosses back above the lower band and back below the upper band.

Pine Script v6
//@version=6
indicator("RSI Settings Tester", overlay = false)

rsiLength = input.int(7, "RSI length", minval = 2)
upperLevel = input.float(80.0, "Overbought level", minval = 50, maxval = 100)
lowerLevel = input.float(20.0, "Oversold level", minval = 0, maxval = 50)

rsiValue = ta.rsi(close, rsiLength)

plot(rsiValue, "RSI", color = color.rgb(37, 99, 235), linewidth = 2)
upperBand = hline(upperLevel, "Overbought", color = color.rgb(220, 38, 38))
middleBand = hline(50, "Midline", color = color.new(color.gray, 40))
lowerBand = hline(lowerLevel, "Oversold", color = color.rgb(22, 163, 74))
fill(upperBand, lowerBand, color = color.new(color.blue, 92))

bullishCross = ta.crossover(rsiValue, lowerLevel)
bearishCross = ta.crossunder(rsiValue, upperLevel)

plotshape(bullishCross, "Cross above oversold", shape.triangleup, location.bottom, color.green, size = size.tiny)
plotshape(bearishCross, "Cross below overbought", shape.triangledown, location.top, color.red, size = size.tiny)

alertcondition(bullishCross, "RSI crossed above oversold", "RSI crossed above the oversold level")
alertcondition(bearishCross, "RSI crossed below overbought", "RSI crossed below the overbought level")
TradingView RSI chart with overbought, midline, and oversold reference levels

The preview uses a standard RSI pane. Your one-minute result will depend on the chart symbol, session, and selected inputs.

Signal rules

Read the crossover, not the label alone

Overbought and oversold describe recent momentum. They do not say that price must reverse on the next bar.

RSI crossover rules and limitations
EventPine conditionWhat it saysWhat it does not say
Cross above oversoldta.crossover(rsi, lower)RSI has moved back above the lower threshold.Price has not necessarily formed a durable bottom.
Cross below overboughtta.crossunder(rsi, upper)RSI has moved back below the upper threshold.Price has not necessarily formed a durable top.
Cross above 50ta.crossover(rsi, 50)Smoothed gains now outweigh smoothed losses.The next price bar is not guaranteed to close higher.

When I review a 1-minute RSI setup, I start with one clear event, such as a cross above 20, and add a trend filter only after the basic rule has been measured. Adding several filters at once makes it hard to tell which condition changed the result.

Testing method

Test a 1-minute RSI without fitting the answer

A short chart interval produces many observations. That makes it easy to find a setting that looks good by chance.

  1. 1

    Write one entry event

    Example: RSI 7 crosses above 20 while price is above the 200 EMA. Define the exit before running the test.

  2. 2

    Use the same data window

    Compare RSI 5, 7, 9, and 14 on the same NQ, QQQ, or SPY dates. Do not change the market halfway through.

  3. 3

    Include trading costs

    Add commission and a slippage assumption. A frequent one-minute rule can lose its apparent edge after costs.

  4. 4

    Reserve unseen dates

    Choose the candidate on one period and check it on a later period that was not used to select the settings.

  5. 5

    Compare stability, not one peak

    Reject a setting when small changes to fees, dates, or thresholds erase the result.

I compare the rule on NQ and QQQ as separate tests. I also keep the default RSI 14 result beside the shorter settings. That baseline makes it obvious when a faster setting only adds more trades without improving the test.

Indicator comparison

Stochastic vs RSI on short charts

Both indicators use a bounded scale, but their inputs differ. Use the one that matches the rule you want to test.

Comparison of RSI and Stochastic oscillator
QuestionRSIStochastic
What is measured?Smoothed gains compared with smoothed losses.Latest close compared with the recent high-low range.
Common default14, with 70 / 3014, 3, 3 with 80 / 20
Typical ruleThreshold, centerline, or divergence event.Percent K and percent D crossover or threshold event.
Main limitationRSI can remain at an extreme during a strong trend.Range-based readings can change quickly on short charts.

For a fuller comparison, read Stochastic Oscillator vs RSI.

FAQ

RSI settings questions

Each answer starts with the conclusion, then states the constraint that matters.

Risk note

This page is an information and testing tool, not investment advice. RSI cannot predict returns. Backtests can overstate a rule when they omit fees, slippage, changing market conditions, or out-of-sample testing.

Turn your RSI rule into Pine Script

Describe the length, levels, filters, and alerts. Pineify writes the TradingView code so you can inspect and test the rule.