7 of the 10 features in our table differ between the DHT22 and the SHT31.
±0.5 °C · ±2 %RH · Single wire (own protocol) · 3.3 to 6 V
A step up from the DHT11: ±0.5 °C, ±2 %RH typical and down to −40 °C, still on one data wire.
Watch out: The humidity error can reach ±5 %RH, one reading every 2 s at most, and its own protocol needs precise timing. About 50 µA between readings.
±0.2 °C · ±2 %RH · I²C · 2.15 to 5.5 V
Sensirion’s well-proven sensor: ±0.2 °C, ±2 %RH, two I²C addresses and a supply from 2.15 to 5.5 V.
Watch out: The SHT4x is newer, smaller and more accurate on humidity for the same job.
| Feature | DHT22 | SHT31 |
|---|---|---|
| Temperature accuracy | ±0.5 °C | ±0.2 °C |
| Temperature range | −40 to 80 °C | −40 to 125 °C |
| Interface | Single wire (own protocol) | I²C |
| I²C address | Not I²C | 0x44, 0x45 (set with a pin) |
| Supply | 3.3 to 6 V | 2.15 to 5.5 V |
| Current between readings | 50 µA | 0.2 µA |
| ESPHome | dht | sht3xd |
Same on both: measures · humidity accuracy · status.
In our table the DHT22 has nothing the SHT31 lacks: choose it for price, availability or a board you already have.
Open the sensor comparison →Tick what your project needs and see which sensor fits best.
SHT31: I²C address, wiring and pull-ups →
The SHT31 offers better temperature accuracy, a wider temperature range, a standard I²C bus shared with other sensors, and less current between readings; in our table the DHT22 has no advantage of its own, apart from price or availability.
Typical datasheet figures: temperature ±0.5 °C for the DHT22 and ±0.2 °C for the SHT31; humidity ±2 %RH and ±2 %RH. Placement matters as much as accuracy: keep either away from the board’s warm parts.
Not as a drop-in: the DHT22 uses a proprietary single-wire protocol and the SHT31 I²C, so the wiring and the code change.
DHT22 vs BME280 DHT11 vs DHT22 AHT20 vs DHT22 SHT31 vs SHT4x AHT20 vs SHT31 DS18B20 vs DHT22
Figures from the makers’ datasheets and product pages and the ESPHome documentation, checked in November 2026.