Twin studies are becoming popular to investigate risk factors for low back pain (LBP) because they consider the genetic factor and allow for more precise estimates of risks. We aimed to identify and summarize the results of studies based on twin samples investigating risk factors for LBP. The MEDLINE, CINAHL, LILACS, Web of Science and EMBASE databases were searched. Prospective and cross-sectional observational studies of LBP involving twins were included. The exposure factors could be genetics (heritability) or environmental such as smoking, alcohol consumption, body mass index and medical history. Pooling was attempted using an inverse variance weighting and fixed effects model. Twenty-seven studies were included. Estimates of heritability effects ranged from 21% to 67%. The genetic component was higher for more chronic and disabling LBP than acute and less disabling LBP. Smoking was significantly associated with LBP [pooled odds ratio (OR) = 3.0; 95% confidence interval (CI) 2.8-3.3] with a longitudinal and a cross-sectional study also identifying a dose-response relationship in people with chronic LBP. Obesity was associated with LBP (pooled OR = 1.9; 95% CI 1.6-2.2) with a cross-sectional study identifying a dose-response relationship. No association between alcohol consumption and LBP was identified. Co-morbidities such as asthma, diabetes and osteoarthritis were associated with LBP (pooled OR ranging from 1.6 to 4.2). The contribution of genetics to LBP appears to be dependent on the severity of the condition. Twin studies could be better used to explore possible causation paths between lifestyle factors, co-morbidities and LBP.