Demography India Vol. 37, Supplement (2008), pp. 77-90 Arvind Pandey*, D. Sahu*, R. M. Mishra*, D. K. Joshi*, S. K. Benara*, U. Sengupta*, Nandini Roy*, R. Adhikary**, D. Pardeshi*** Integrated Behavioural and Biological Assessment—National Highways: Survey Methodology and Implementation Introduction S URVEILLANCE of mobile, hard-to-reach and ‘hidden’ populations is crucial for containing the HIV spread in low prevalence settings and slowing the rate of transmission in high prevalence settings. However, conventional facility and community based data collection strategies have been ineffective for a number of reasons, e.g. there are subpopulations whose behaviours are considered illegal or illicit by general population (Magnani et al., 2005) and those who are mobile or staying away from home for a number of days and indulging in high-risk behaviour. The primary challenge has been to obtain representative samples of them (Mills et al., 1998; Schwartlander et al., 2001). Having addressed these issues, a number of alternative survey designs have been postulated. Perhaps the most commonly used method is snowball sampling that entails identifying an initial number of members from whom the desired data are gathered and who then serve as seeds (Heckathorn, 1997; Kish, 1995; Thompson and Collins, 2002, Villarinho et al., 2002). The seeds are subsequently taken to identify other subgroup members who engage in the same types of behaviours and include in the sample. The individuals in turn are asked to provide information on other subgroup members. The process continues until either a target sample size is achieved or the sample is ‘saturated’. A similar technique, known as targeted sampling, extends the ideas of snowball sampling to include an initial ethnographic assessment aimed at identifying the various networks or subgroups that might exist in a given setting (Watters and Biernacki, 1989; Heckathorn, 1997). The subgroups so identified are then treated as sampling strata and samples are chosen within each stratum using systematic sampling. Respondent driven *National Institute of Medical Statistics, ICMR, New Delhi, India. E-mail: arvindpandey@icmr.org.net **Family Health International, New Delhi, India. ***National AIDS Research Institute, ICMR, Pune, India.