We reply to the commentaries on the lead papers by Neale et al. [1994a] and Duffy and Martin . Topics covered include power calculations, cross-sectional measurement vs. lifetime reports, the appropriateness of the direction of causation (DOC) model, extensions to study causation between the latent variables, sampling of subjects, and heterogeneity. We consider the potential of combining genetically informative research designs with multivariate longitudinal and experimental methods. © 1994 Wiley-Liss, Inc.