How to see if a value crossed over or under another in TradingView Pine?

TL;DR
Call ta.crossover(seriesA, seriesB) for upward crosses, ta.crossunder(seriesA, seriesB) for downward crosses, and ta.cross(seriesA, seriesB) when you only need to know a cross happened.

At a Glance


Difficulty	Beginner
Time to implement	5-10 min
Category	Signals

Quick Actions

Quick Start

//@version=5
indicator("Crossover demo", overlay=false)

fast = ta.ema(close, 12)
slow = ta.ema(close, 26)

longSignal  = ta.crossover(fast, slow)
shortSignal = ta.crossunder(fast, slow)

Tip. `ta.crossover` returns `true` only on the bar where the crossover occurs. On subsequent bars it reverts to `false` unless another cross happens.

Why It Matters

Crossovers are the backbone of countless strategies—moving averages, oscillator thresholds, price vs. indicators. Pine’s ta.cross* functions encapsulate the logic, saving you from manual state tracking and reducing off-by-one errors.

What You’ll Learn

Understand the difference between crossover, crossunder, and generic cross detection.
Use cross signals for entries, exits, and alert conditions.
Guard against false positives with filters (e.g., minimum separation).
Visualise cross points with shapes and labels.

Quick Reference

Call	Purpose
`ta.crossover(series1, series2)`	Returns `true` when `series1` crosses from below to above `series2`.
`ta.crossunder(series1, series2)`	Returns `true` when `series1` crosses from above to below `series2`.
`ta.cross(series1, series2)`	Returns `true` for any cross (either direction).
`plotshape(condition, ...)`	Highlights the cross on chart.
`alertcondition(condition, title, message)`	Triggers alerts when a cross occurs.

Implementation Blueprint

Create the series to compare
These can be moving averages, price vs. indicator, or any custom calculation.
```
basis = ta.sma(close, 50)
trigger = ta.ema(close, 10)
```

Detect the cross
Choose the appropriate helper for the direction you care about.

bullish = ta.crossover(trigger, basis)
bearish = ta.crossunder(trigger, basis)

Act on the signal
Plot shapes, fire alerts, or execute strategy orders when the cross occurs.

if bullish
    strategy.entry("Long", strategy.long)
plotshape(bullish, title="Bull cross", location=location.belowbar, shape=shape.triangleup, color=color.new(color.green, 0))

Example Playbook

//@version=5
indicator("EMA crossover highlighter", overlay=true, max_labels_count=500)

fastLen = input.int(21, "Fast EMA")
slowLen = input.int(55, "Slow EMA")

fast = ta.ema(close, fastLen)
slow = ta.ema(close, slowLen)

bullish = ta.crossover(fast, slow)
bearish = ta.crossunder(fast, slow)

plot(fast, "Fast", color=color.new(color.green, 0))
plot(slow, "Slow", color=color.new(color.red, 0))

plotshape(bullish, title="Bullish cross", location=location.belowbar, shape=shape.arrowup, color=color.new(color.green, 0), text="Bull")
plotshape(bearish, title="Bearish cross", location=location.abovebar, shape=shape.arrowdown, color=color.new(color.red, 0), text="Bear")

alertcondition(bullish, "Bullish crossover", "Fast EMA crossed above slow EMA")
alertcondition(bearish, "Bearish crossunder", "Fast EMA crossed below slow EMA")

Why this works.

Separate variables for bullish and bearish crosses make downstream logic explicit.
Shapes and alerts reuse the same booleans, keeping visuals and notifications aligned.
Input lengths let traders adjust sensitivity without modifying code.

Pro Tips & Pitfalls

Filter out rapid-fire crosses by requiring minimum bar separation or additional conditions (e.g., trend direction).
Combine with ta.barssince to measure time since the last cross.
For threshold-based crosses (e.g., RSI crossing 50), compare the oscillator against a constant (ta.crossover(rsi, 50)).
Remember that crosses on higher timeframes may repaint slightly until the higher-timeframe bar closes.

Troubleshooting & FAQ

Why do I see multiple true values in a row?

ta.cross* returns true once per bar. If you see repeated signals, ensure you’re not storing the boolean in a var and failing to reset it.

The cross fires even when the series just touches

By definition, a cross triggers when series1 moves from one side to the other—even if it only touches for a split second. Add additional conditions (e.g., series1 > series2 + buffer) to require more separation.

Can I detect crosses on historical bars only?

Run the logic as-is; it already evaluates on historical bars. Just be mindful that on real-time bars the signal can flip until the bar closes.

Key Takeaways

ta.crossover, ta.crossunder, and ta.cross provide the simplest way to detect directional crosses.
Reuse the same boolean for plotting, alerts, and trade logic to keep behaviour consistent.
Add filters when you want to reduce noise or confirm crosses with additional criteria.
Cross detection is foundational—mastering it unlocks dozens of classic Pine strategies.

Keep Going

Open AI Editor to apply the pattern with guided assistance
Browse Strategy Examples for ready-to-run templates
Connect to Live Trading when you are ready to automate

Adapted from tradingcode.net, optimised for Algo Trade Analytics users.