The District Level Household Survey (DLHS) was initiated in 1997 with a view to assess the utilization of services provided by government health care facilities and people’s perceptions about the quality of services. The District Level household Survey (DLHS -3) is the third in the series of district surveys, preceded by DLHS-1 in 1998-99 and DLHS-2 in 2002-04. As in DLHS-3, the International Institute for Population Sciences (IIPS) was the nodal agency to conduct DLHS-1 and DLHS-2.  DLHS-3, like other two earlier rounds, is designed to provide estimates on important indicators on maternal and child health, family planning and other reproductive health services. In addition, DLHS-3 provides information on important interventions of National Rural Health Mission (NRHM). Unlike other two rounds in which only currently married women age 15-44 years were interviewed, DLHS -3 interviewed ever-married women (age 15-49) and never married women (age 15-24).

The sample size among the districts in the country varies according to their performance in terms of ante-natal care (ANC), institutional delivery, immunization, etc. and it was fixed based on information related to such indicators from DLHS-2. For low performing districts, 1500 Households (HHs), for medium performing districts, 1200 HHs and for good performing districts, 1000 HHs were fixed as sample size.  In case of Belgaum, sample size was 1200 households with 10% additional HHs to take care of non-response/refusal, etc. 

The survey used two-stage stratified random sampling in rural and three-stage stratified sampling in urban areas of each district. The information from 2001 Census was used as sampling frame for selecting primary sampling units (PSUs). In rural areas, all the villages in the district were stratified into different strata based on population /HH size, percentage of SC/ ST population, female literacy (7+), etc. The required number of villages from each strata were selected with probability proportional to size (PPS). In selected primary sampling units (villages), household listing was done and required number of households were selected using systematic random sampling.
For larger villages (more than 300 HHs) segmentation was carried out.  In case of 300 to 600 HHs, two segments of equal size were made and one was selected using PPS. For PSUs having more than 600 HHs, segments of 150 HHs were created depending on the size and then two segments were selected using PPS.  In case of urban areas, number of wards were selected using PPS at first stage.  In a selected ward, one enumeration block from 2001 census was selected again using PPS. Procedure for segmentation, houshold selection, etc, was same as in the case of rural PSUs.  

The uniform bilingual questionnaires, both in English and in local language, were used in  DLHS-3 viz., Household, Ever Married Women (age 15-49), Unmarried Women (age 15-24), Village and Health facility questionnaires.

In the household questionnaire, information on all  members of the household and the socio-economic characteristics of the household, assets possessed, number of marriages to usual members of the household since January 2004 and deaths in the household since January 2004 etc. was collected.  In case of female deaths, attempts were made to assess maternal death.  The household questionnaire also collected information on respondent’s knowledge (seen/read/heard) about messages related to various government health programmes being spread through  media and other sources. 

The ever married women's questionnaire consisted of sections on women's characteristics, maternal care, immunization and child care, contraception and fertility preferences, reproductive health including knowledge about HIV/AIDS.

The unmarried women's questionnaire contained information on her characteristics, family life education and age at marriage, reproductive health-knowledge and awareness about contraception, HIV / AIDS, etc.

The village questionnaire contained information on availability of health,  education and other facilities in the village and whether the health facilities are accessible throughout the year.