Diagnosis of sciatica mainly relies on pain reproduction by stretching of the lumbar roots since neurological examination and medical history are usually not sufficient to guarantee diagnosis. The Lasègue test is the most popular method, which starts with the straight leg-raising test (SLR). However it is not perfect, and is not always well performed or interpreted. Passive ankle dorsiflexion at the end of the SLR (Bragard test) is more sensitive, but can also remain normal in some cases of sciatica. Other stretching tests can help to recognise lumbar root damage in patients with poorly defined pain in a lower extremity: firstly, the Christodoulides test, i.e. reproduction of L5 sciatic pain by a femoral stretch test; secondly, the Slump test, performed on a patient in a sitting position, by slowly extending their painful leg then passively bending their neck (or the opposite); and thirdly, the Bowstring test, which requires, at the end of the Lasègue test, once the knee has been slightly flexed, pressing on the course of the peroneal and/or tibial nerves in the popliteal fossea to try and reproduce the exact pain felt by the patient. The combination of all these tests takes less than 2minutes, and could improve both the sensitivity and specificity of the physical examination for the diagnosis of sciatica. This article is a review of the limitations of the Lasègue/SLR tests and of the efficacy of these other tests for stretching the lumbar roots.