This report provides an in-depth description of the first wave of household data collected for a randomised field experiment to measure the impact of microcredit on poverty reduction among poor rural women in Mongolia. The experiment consists of two 'treatments': a group lending product with group responsibility (so-called 'joint liability') and an individual loan product. In a previous report (Attanasio et al, 2008) we detailed the background information to the project, the partner institutions involved, the randomisation methodology, loan products, and outcome variables of interest. In this report, we analyse the data collected from the first wave (the baseline). We provide descriptive statistics relating to our sample along a wide range of dimensions such as education choices, assets, savings, debt, income, enterprises, consumption and transfers. The analysis of this population is of interest in its own right and gives a first snapshot of the target population which is not available from existing data sources. We show formal comparisons of these characteristics between treatment and control groups, an important test of how well the randomisation actually worked, and thus a crucial pre-requisite against which the program will be evaluated in around a year's time.