The chapters in this volume bring together important recent advances in the areas of (i) econometrics of panel data, (ii) limited dependent variable models, and (iii) limited dependent variable models with panel data. Panel data offers researchers many more possibilities than pure cross-sectional or time series data. Like cross-sectional data, panel data describes each of a number of individuals. Like time series data, it describes changes through time. By blending characteristics of both cross-sectional and time series data, panel data can be used, for example, to (i) expand sample size, (ii) allow the specification of more complicated behavioral hypotheses, (iii) capture cross-sectional variation and dynamic behavior, (iv) lessen the problem of multicollinearity, (v) provide possibilities for reducing omitted variable and estimation biases, (vi) improve accuracy of parameter estimates, (vii) obtain more accurate prediction of individual outcomes. However, the analysis of panel data also raises a number of new issues. For instance, in the case of short dynamic panel data models with large cross-section units, it is known that dealing with the initial values and incidental parameters problem can be complex. In other applications, such as non-linear panel data models with fixed effects, a general solution to the problem may not exist. A notable example is estimation of probit models with fixed effects. Also panel data based on economic surveys are very often qualitative in nature, and have limited variations due to self-selection, and truncation.