Our overall score consist of: Performance (80%) and Ease of use (20%).

Based on comparative speed of response under similar smoke levels in a range of tests. Made up of 'real fire' scenarios 70% (made up of smouldering foam 40%, burning wood 20%, smouldering wood 20%, and burning oil 20%) and nuisance smoke scenarios 30%.

Based on the ease of installing the battery, installing the alarm, and the supplied instructions, weighted equally.

Based on the comparative speed of response to smouldering polyurethane foam. The faster the alarm detects this type of fire, the higher the score.

Based on the comparative speed of response to smouldering wood. The faster the alarm detects this type of fire, the higher the score.

Based on the comparative speed of response to burning (flaming) wood. The faster the alarm detects this type of fire, the higher the score.

Based on the comparative speed of response to burning (flaming) cooking oil. The faster the alarm detects this type of fire, the higher the score.

Based on the comparative speed of response to nuisance sources ('false alarms') from cooking toast, smouldering cooking oil and cooking a hamburger patty. The faster the alarm detects this type of fire, the lower the score. Models with high scores are less likely to go off due to nuisance smoke/fumes, such as from cooking.

We recommend models that score at least 70% overall. The score is based on how well they perform compared to the others on test. We also set a minimum score for the smouldering foam test: photoelectric alarms must score at least 80% in this test, as this is exactly the sort of smoke they should be quick to detect. Ionisation and dual sensor models must score at least 70% in the smouldering foam test. Ionisation models are less effective at detecting this sort of smoke, and for dual sensor models, fitting both types of detector into one unit may involve a compromise in the internal space; but in either case they should be at least good at detecting this common fire type.

'Tested' models are currently available. 'Discontinued' models are no longer available, but may still be included in the review so you can see how they compare.

The manufacturer's recommendation for the best rooms to install this alarm.

Regular 9V or AA batteries need to be replaced annually. 10-year lithium batteries last the life of the smoke alarm and don't need to be replaced. Mains-powered units have a battery back-up in case of power outages.

The warranty period (in years). Alarms should last for 10 years so a 10-year warranty is best.

The stated country of manufacture.

The manufacturer or distributor. Note that some distributors don't list all their brands and models on their site.

Silences the alarm temporarily. Useful in case of a nuisance alarm.

With an interconnectable model, you can connect multiple alarms (usually by wire) so that when one goes off, all go off.