Varicella infection during pregnancy has serious and/or difficult implications and in some cases lethal outcome. Though epidemiological studies in developing countries reveal that a significant proportion of patients may remain susceptible during pregnancy, such an estimate of susceptible women is not known in India. We designed this study to study the prevalence and factors associated with susceptibility to varicella among rural and urban pregnant women in South India. We prospectively recruited 430 pregnant women and analysed their serum varicella IgG antibodies as surrogates for protection. We estimated seroprevalence, the validity of self-reported history of chickenpox and factors associated with varicella susceptibility. We found 23 (95% CI 19.1–27.3) of women were susceptible. Nearly a quarter (22.2%) of the susceptible women had a history of exposure to chickenpox anytime in the past or during the current pregnancy. Self-reported history of varicella had a positive predictive value of 82.4%. Negative history of chickenpox (adjusted prevalence ratio (PR) 1.85, 95% CI 1.15–3.0) and receiving antenatal care from a rural secondary hospital (adjusted PR 4.08, 95% CI 2.1–7.65) were significantly associated with susceptibility. We conclude that high varicella susceptibility rates during pregnancy were noted and self-reported history of varicella may not be a reliable surrogate for protection.