The evidence base for changes in electromyographic (EMG) onset and intensity of vastus medialis (VM) and vastus lateralis (VL) in patients with patellar instability is systematically reviewed. The databases AMED, CINAHL, the Cochrane Library, EMBASE, Ovid Medline, Physiotherapy Evidence Database (PEDro), Pubmed and Zetoc were searched from their inception to March 2008, in addition to a manual search of relevant journals. All English-language papers assessing the relative EMG onset and intensity of vastus medialis oblique (VMO) and VL, in patients with patellar dislocation, subluxation and general instability were included. Five papers, consisting of 73 knees with patellar instability, were reviewed. No studies were identified assessing VMO–VL onset in patients with patellar instability. Four studies reported no difference in relative EMG intensity of VMO and VL in patients with patellar instability compared with asymptomatic control subjects. One study reported some evidence of a difference in VM to VL EMG intensity in one cohort of patients with patellar subluxation. The CASP appraisal of the evidence base highlighted a number of methodological weaknesses. There was no robust evidence for any difference in the relative intensity of EMG activity between the VMO and VL in patients with patellar instability. There is no good-quality research evidence to suggest that abnormal vastii EMG intensity or onset are aetiological factors associated with patellar instability.