OBJECTIVE To summarise recent evidence from case reports (published January 1995-September 2001) of adverse events after cervical spine manipulation. DATA SOURCES Five computerised literature searches (MEDLINE-Pubmed; EMBASE, the Cochrane Library, AMED [Allied and Complementary Medicine Database], and CISCOM [Centralised Information Service for Complementary Medicine] were performed. No language restrictions were applied. STUDY SELECTION All case reports containing original data of adverse events after cervical spine manipulation were included. DATA EXTRACTION All articles were evaluated and key data extracted according to pre-defined criteria: patient's age, sex and diagnosis; type of therapist; type of treatment; nature of adverse event; method of diagnosis; and clinical outcome. DATA SYNTHESIS Thirty-one case reports (42 individual cases) were found. The patients were equally distributed between the sexes (21 male, 20 female, one unknown) and mostly middle-aged (range, 3 months to 87 years). Most were treated by chiropractors. Arterial dissection causing stroke was reported in at least 18 cases. CONCLUSIONS Serious adverse events after cervical spine manipulation continue to be reported. As the incidence of these events is unknown, large and rigorous prospective studies of cervical spine manipulation are needed to accurately define the risks.